research visual abstract

From article to art: Creating visual abstracts - Parts 1 & 2: A Guide to Visual Abstracts

Michelle Feng He

Ginny Pittman

About this video

What is a visual abstract and why should you use visual abstracts in your research? How can you create a visual abstract and what message from your research should you select? Hear from Michelle Feng He and Ginny Pittman, as they guide you through visual abstracts in parts 1 & 2 of our "From article to art: Creating visual abstracts" module.

About the presenters

Publisher, Elsevier

Based in New York, Michelle joined Elsevier from Springer Nature where she developed the journal strategies across their oncology, surgery, pathology, and life-sciences programs. Her interests and strengths lie in data science, society partnership management, and reviewer engagement programs. As a Publisher, Michelle’s first priority is ensuring that all journal stakeholders have the necessary data to make strategic decisions regarding the impact and growth of their journals.

Executive Publisher, Elsevier

Ginny leads the orthopedics portfolio in the US, with 18 open access and subscription journals. She has over 25 years of experience in scholarly research publishing and educational product development, making an impact in the healthcare and life sciences fields at Wolters Kluwer, Mary Ann Liebert Publishers and Wiley-Blackwell. She has presented to author communities at medical institutions globally and has created new approaches to portfolio development, metrics, and author/editor training and tools, earning multiple awards for innovation.  

Visual abstract design resources

How to make a graphical abstract?

Table of Contents

Scientific research and publishing have been evolving rapidly. As the sheer volume of published papers continues to grow exponentially, graphical abstracts have emerged as one of the most important aspects of communicating research effectively and improving the visibility and accessibility of findings. Academic journals are increasingly requesting the submission of graphical abstracts along with text-based abstracts. This stems from the wide-ranging benefits offered by graphical abstracts.  

What is a graphical abstract?  

  A graphical abstract is a visual representation of a study’s key findings. It distils complex information into easily understandable, visually appealing formats that enhance comprehension and retention.

Graphical abstracts can take the form of graphs, illustrations, diagrams, images, or photographs. They are typically a single image that combines text, symbols, and visual elements to convey the key highlights of the study clearly and accurately.  

  An effective graphical abstract captures readers’ attention and directs them to the key aspects of the study without forcing them to read the entire paper. Usually presented at the beginning of a research paper, graphical abstracts provide an effective alternative to text—one that can engage both specialists and non-specialists alike.

Features of a good quality graphical abstract  

Graphical abstracts are essential for conveying research essence effectively with minimal text and maximum visuals. Here are the basic features of a good-quality graphic abstract:

• A simple and clear presentation is crucial for a broad audience’s understanding
• Avoid overly bright colors, fancy fonts, and distracting design elements
• Use relevant images that communicate the central message clearly
• Adhere to technical specifications for proper layout, format, and high resolution

How do you make a graphical abstract for your research paper?  

The following key steps will guide you in preparing a powerful graphical abstract. 

• Conceptualize: Identify key findings or aspects of the study. As a researcher, you may be tempted to include all the main results and conclusions. This, however, will make it unwieldy. Make sure to focus only on core findings that can effectively communicate the critical points of your study.  
• Outline: Once the key messages have been identified, prepare a rough outline or sketch of the graphical abstract. Plan the layout, including the placement of the images or illustrations and the textual elements. Ensure that the arrangement of the images and text is orderly so that the narrative is clear. Avoid using too many elements or excessive text, as it can confuse readers. 
• Design: Use images and visuals appropriate for the message being communicated. The color scheme and fonts selected should also be easy on the eyes.  

Example of a graphical abstract  

Source: Graphical Abstract Examples — Sage Research Methods Community (sagepub.com) 

Five things to consider for an effective graphical abstract  

• Know your audience: While designing graphical abstracts, keep in mind the nature of your audience. Would it be a specialized audience with knowledge about the subject or a general one?  
• Focus on the key messages: Communicate two or three key findings or results. Avoid presenting too much information, as it can look cluttered and be difficult to interpret. 
• Use the right tools: It is essential to ensure that the visual elements are relevant to the interest of the readers. To make your graphical abstract impactful and attractive, you can always use tools such as Mind the Graph . With a library of 75,000+ scientific illustrations, 300+ customizable templates, and the availability of multiple formats, Mind the Graph can help researchers and students save time and come up with great graphical abstracts.   
• File Format: Once created, remember to save the graphical abstract in a high-resolution format (e.g., TIFF, PNG) to ensure optimal quality for publication and make it easier to share online. 
• Seek feedback: This will help you understand whether the key messages are being clearly communicated and whether the visual elements are appealing and interesting. 

While creating a graphical abstract is often seen as a challenging task, following the steps outlined in this article will enable you to create impactful visuals that can enhance the impact of your research. Given the speed at which digitalization is moving, the need for graphical abstracts will only continue to grow.

Researchers can use the many online tools available to aid them in their efforts to develop their visual communication skills. Do keep in mind, however, that while using the plethora of online tools can help streamline the design process, it is important to ensure the right mix of AI and human creativity to ensure accuracy and originality. 

Paperpal is a comprehensive AI writing toolkit that helps students and researchers achieve 2x the writing in half the time. It leverages 22+ years of STM experience and insights from millions of research articles to provide in-depth academic writing, language editing, and submission readiness support to help you write better, faster.  

Get accurate academic translations, rewriting support, grammar checks, vocabulary suggestions, and generative AI assistance that delivers human precision at machine speed. Try for free or upgrade to Paperpal Prime starting at US$19 a month to access premium features, including consistency, plagiarism, and 30+ submission readiness checks to help you succeed.  

Experience the future of academic writing – Sign up to Paperpal and start writing for free!  

Related Reads:

• How to Make Your Thesis Supervision Work for You
• How to Write a Scientific Paper in 10 Steps 
• How to Avoid Plagiarism When Using Generative AI Tools
• 7 Best Referencing Tools and Citation Management Software for Academic Writing

Introducing AI Review, Paperpal’s New Virtual Research Coach for Busy Academics

Innovation and technology in peer review, you may also like, academic integrity vs academic dishonesty: types & examples, the ai revolution: authors’ role in upholding academic..., the future of academia: how ai tools are..., how to write your research paper in apa..., five things authors need to know when using..., 7 best referencing tools and citation management software..., what’s the best chatgpt alternative for academic writing, addressing peer review feedback and mastering manuscript revisions..., what is the importance of a concept paper..., how to write a high-quality conference paper.

Loading metrics

Open Access

Perspective

Ten simple rules for designing graphical abstracts

Roles Conceptualization, Funding acquisition, Project administration, Visualization, Writing – original draft, Writing – review & editing

* E-mail: [email protected] (HKJ); Martin.bornhä[email protected] (MB)

Current address: Centre for Data Analysis, Visualisation and Simulation, University of Applied Sciences of the Grisons, Chur, Switzerland

Affiliation National Center for Tumor Diseases—University Cancer Center (NCT-UCC), Universitätsklinikum Carl Gustav Carus an der Technischen Universität Dresden, Germany

Roles Funding acquisition, Visualization, Writing – review & editing

Affiliations National Center for Tumor Diseases—University Cancer Center (NCT-UCC), Universitätsklinikum Carl Gustav Carus an der Technischen Universität Dresden, Germany, Medical Clinic 1, Universitätsklinikum Carl Gustav Carus an der Technischen Universität Dresden, Germany

• Helena Klara Jambor, 
• Martin Bornhäuser

Published: February 1, 2024

• https://doi.org/10.1371/journal.pcbi.1011789
• Reader Comments

Citation: Jambor HK, Bornhäuser M (2024) Ten simple rules for designing graphical abstracts. PLoS Comput Biol 20(2): e1011789. https://doi.org/10.1371/journal.pcbi.1011789

Editor: Scott Markel, Dassault Systemes BIOVIA, UNITED STATES

Copyright: © 2024 Jambor, Bornhäuser. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Funding: HKJ received a salary from an habilitation award of the Medical Faculty of the Technische Universität Dresden. HKJ and MB received project funding from the Hochschulstiftung Medizin Dresden. MB received funding from the MSNZ program of the Deutsche Krebshilfe. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Explanatory graphics that summarize knowledge are common in science communication. These graphics integrate new insights with the existing body of knowledge in a particular field of research. Explanatory graphics have been widely used in chemistry for many years to depict structures [ 1 ], and they have now gained popularity across various scientific disciplines as graphical abstracts [ 2 ]. Scientific journals are increasingly asking authors to provide graphical abstracts along with a paper to attract audiences online and on social media. These graphical abstracts are prominently displayed on the journals websites, embellishing the table of contents, and serving as a visual pendant to the written abstract. Due to this usage, graphical abstracts are also referred to as “TOC” image or “thumbnail views.”

Graphical abstracts are not intended to provide a complete understanding of a research article, even though they are often presented online with just the title of the work. A study confirmed graphical abstracts by themselves are insufficient to comprehend the key message of a paper [ 3 ]. Instead these visuals serve to attract attention and are meant to be read in conjunction with the written abstract. According to Cell press guidelines, graphical abstracts should inspire audiences to browse, stimulate their interdisciplinary curiosity, and allow them to rapidly screen for papers in journals [ 4 ]. As graphical abstracts are a relatively recent addition to the publishing landscape, quantitative data on their usage and usefulness are still limited. However, early analyses indicate that while graphical abstracts do not necessarily increase full-text reads or citations, they do enhance the abstract views [ 5 ] and boost altimetric attention scores of articles [ 6 ].

Like other explanatory visualizations, graphical abstracts have common features such as a central visual element, often icons, diagrams or photos, explanatory text, and use clear layout and color schemes to increase readability. These elements are often structured using arrows and lines and enhanced with color. The design elements of graphical abstracts were recently quantified in a research study that classified graphical abstracts based on their overall organization [ 2 ]. In their work, Hullman and Bach revealed the diversity of graphical abstracts in the current literature, and in particular, the many possibilities to use layout for readability. They also pinpointed common problems associated with graphical abstracts, such as inconsistent visual styles, unclear relationships between pictures, and missing annotations. These challenges were also identified in a complementary qualitative study of graphical abstracts [ 7 ].

Training of scientists, especially early career researchers, in the art of crafting comprehensible and attractive graphical abstract has been somewhat lagging. A brief guide for graphical abstract design is available for medical writers [ 7 ] and for creating overview figures [ 8 ]. However, most scientist are not trained in data visualizations or visual communication [ 9 , 10 ], and even less so in creating explanatory visuals of their research. It’s important to note that visual design is a nontrivial endeavor. Publishing houses, journals, and major research institutes often employ visual teams to create attractive explanatory figures for scientific data.

Here, we present 10 simple rules for designing graphical abstracts. The 10 rules are informed by our experience teaching biologists, clinicians, students, and established scientists, and from jointly preparing graphical abstracts for publications and grants ( Fig 1 ). The article discusses all aspects of graphical abstract preparation, from foundational decisions about the message and the key visuals (1 to 3), to designing the layout (4 to 6), and complementing the design with text (7) and color (8). We also provide an overview of tools and software commonly used for making graphical abstracts (9) and highlight the benefit of feedback in the process (10). The order of the 10 rules reflects our “design pipeline” from starting with a draft to implementing the draft electronically; however, as with all creative processes, you are encouraged to adapt the process to your own style.

• PPT PowerPoint slide
• PNG larger image
• TIFF original image

The evolution of a graphical abstract, from sketch (A) to a rapid Biorender draft (B) and final implementation in a graphical software program (C). All drawings by HKJ, licensed under CC0, https://doi.org/10.6084/m9.figshare.24486061.v1 .

https://doi.org/10.1371/journal.pcbi.1011789.g001

Rule 1: Key message for audience

Before embarking on the design of a graphical abstract, it is essential to know your message. This tip is not specific to graphical abstracts, but also essential for producing an understandable and clear visualization. The process of defining a key message varies. Some start with doodling on a post-it, some with key visuals, and some by iteratively shortening the abstract to 1 or 2 punchy sentences. Ideally already at this stage co-authors are involved and provide feedback (see #10). Recent tools, such as chatGPT, may be helpful in facilitating a dynamic exchange and the concise distillation of the core message. Whichever route is yours, without a clear central message, it will be impossible to design a clear graphical abstract and reach the goal of visually summarizing your research paper.

Rule 2: Pictures and pictograms

The key components of every graphical abstract are the visual elements. Most often, graphical abstracts include pictograms or symbols and, less commonly, iconic microscope ( Fig 1 ) or photographic images, or data (see #5). Pictograms may also be hand-drawn, but mostly biologists use simple shapes, circles, ellipses, and rectangles, when creating pictograms from scratch. In recent years, numerous icon collections have become available, many of which are free to re-use and do not always need attribution. In most icon repositories pictograms can be downloaded as PNG (Portable Network Graphic), a raster-graphics format for lossless data compression. PNGs are ready to use in graphic software but not adaptable. Alternatively, icons are provided as SVG (Scalable Vector Graphics), an image format that can also be used interactively on the web and is fully adaptable in appearance with graphic software.

For general icons, many repositories exist for simple icons:

• PowerPoint offers inbuild image and icon libraries and many pre-drawn shapes that are free to use.
• Fontawesome ( https://fontawesome.com ) is a Unicode-based icon library that can be installed locally as a font for graphic programs, downloaded as full icon library, or downloaded as individual SVG images.
• Nounproject ( https://thenounproject.com ) is a large repository sourcing icons from various designers. Hence, the available icons are vast, but also not matched in style. These icons can be used for free with attribution as SVG or PNG.
• SVGrepo ( https://www.svgrepo.com ) is the largest SVG icon library, which additionally provides search functions for icon style and appearances such as color, rounded or sharp icons.

Biology and Medicine require specific icons which are available in the following repositories:

• Phylopic ( https://www.phylopic.org/ ) offers shapes of numerous animals, plants, and further model organisms, e.g., for phylogenetic trees.
• The EBI ( https://www.ebi.ac.uk/style-lab/general/fonts/v1.3/ ) provides some general scientific icons.
• Reactome ( https://reactome.org/icon-lib ) provides scientific pictograms and chemical drawings for free re-use and encourages the upload of user-designed pictograms for sharing with the scientific community.
• Smart ( https://smart.servier.com/ ) is a free collection of medical drawings from Servier Medical Art and can be downloaded as a full slide-deck and used with attribution.
• Bioicons ( https://bioicons.com/ ) is an expanding set of biology and laboratory icons from Petri dishes to model organisms available under free licenses (CC0). Initially by Servier, the drawings are expanded with user provided samples.
• Health Icons ( https://healthicons.org/ ) is a global volunteer effort to create common icons for many specialized medical scenarios available under creative commons license (CC0).

In a graphical abstract, all icons should have a similar overall appearance, meaning the same line-width, color scheme, and level of detail. Icons from the same source and/or the same designer usually have such a similar look. Fig 2A and 2B shows 2 versions of a graphical abstract with a poor and improved icon combination. If icons from several sources are combined, you could match their style by adapting the SVG-pictograms in graphic software.

(A) All pictograms used have similar overall appearance (color, size, design, modified from [ 11 ]). (B) Poor combination of pictogram for the same workflow: pictograms have different overall appearance. Icons in A: Fontawesome, Fonticons, Inc. Icons in B: Microscope: Bioicons DBCLS https://togotv.dbcls.jp/en/pics.html is licensed under CC-BY 4.0; Laptop: Icon by Simon Dürr https://twitter.com/simonduerr is licensed under CC0 https://creativecommons.org/publicdomain/zero/1.0/ ; Image/slider: see A; Colors and people: drawn by HKJ; Newspaper: https://www.svgrepo.com/svg/301104/newspaper-news , CC0.

https://doi.org/10.1371/journal.pcbi.1011789.g002

For inspiration, you may wish to explore one of the earliest icon libraries, the ISOTYPE. The ISOTYPE system was developed by Otto Neurath in the 1920ies in Vienna as a visual communication tool for low-literate populations. The designs are from Gerd Arntz and were later continued by Marie Neurath ( http://gerdarntz.org/isotype.html ).

Rule 3: Data and charts as key visual

At times pictograms cannot sufficiently represent a key message. You then may wish to include data or charts in your graphical abstracts. When your data are medical, microscopy, or photo images they may be self-explanatory in graphical abstracts. When you want to instead include data plots, you should aim for chart types that are understandable even in the short view time of graphical abstracts. Most of us are familiar with bar charts, which are the most common chart type in scientific publications [ 12 , 13 ], and with pie and line charts, plot types we usually learn in school ( Fig 3 ). These charts employ core principles of visual perception: in bar charts we almost intuitively compare lengths, in pie charts the slice areas, and in line charts we look for up- or downward trends [ 14 ].

Note that the core message (increases, is most, one third…) is communicated without axis details, labels, and legends.

https://doi.org/10.1371/journal.pcbi.1011789.g003

When it is necessary to signify the use of a specific method in graphical abstracts, sometimes method-specific charts are employed as visual placeholders. For example, t-SNE plots may represent single-cell data, red/green heatmaps can denote microarray data, and circoid plots are indicative of genomic approaches. However, it’s important to note that readers of your graphical abstract are unlikely to delve into the details in these advanced graphics. In such instances, a simple version of that charts should be used, featuring only a handful of data points or categories. Details like tick marks, axis label, and legends can be omitted. For a comprehensive understanding of different chart types and their appropriate use a valuable resources is the Data Visualization Catalogue ( https://datavizcatalogue.com/ ).

Rule 4: Layout: The dimensions

Layout describes the organization of visual elements on the page ( Fig 4 ). First, consider the space that you have available to fill. A graphical abstract for a journal website is typically shown as a square and rarely in rectangle format ( Fig 4A ). On many websites and applications, the graphical abstract has a final size not much larger than a postage stamp. When a graphical abstract is the first figure of an article, poster, or grant application, you may also opt for a landscape rectangle format. Whenever choosing a layout, you should consider how to fill the area best. In grant applications space is very limited, filling the entire width of a line may then be a best choice to not waste precious space ( Fig 4B ).

Different dimensions (A) and how they merge with text on a page (B).

https://doi.org/10.1371/journal.pcbi.1011789.g004

Rule 5: Layout: Reading direction

The layout should provide a clear entry point into your graphical abstract. Typically, we read from left to right, and top to bottom in either columns or rows. You should therefore arrange all elements of the graphical abstract along your chosen reading direction [ 15 ].

For depiction of linear processes that have a clear beginning and end, an organization from left to right is most suitable: time is usually shown as the independent variable on the x-axis in graphs. Linear processes are workflows, experimental pipelines, embryonic development, cellular differentiation, or disease progression. Alternatively, you can consider a circular layout for cyclic events such as daily or annual events, metabolic cycles, or processes like cell division. For static observations, e.g., contrasting 2 scenarios or providing 2 levels of details for 1 scenario, you could consider 2 parallel or nested organization [ 2 ]. Fig 5 summarizes the most common organizational layouts of graphical abstracts.

https://doi.org/10.1371/journal.pcbi.1011789.g005

Rule 6: Connecting the elements: Arrows and arrangement

Arrows are a key element for all explanatory graphs and visual abstracts. With arrows, we connect text, pictograms, images, and charts into a sequential narrative or “storyline” and consequently they are the most common graphical element in explanatory life science figures [ 16 ]. Arrows can reinforce your chosen reading direction but arrows can also signal any exception from this reading direction. A clear layout supported by arrows helps to quickly orient your audiences in a visualization.

In graphical abstracts, arrows have several distinct appearances and also distinct functions. Arrows also include arrowheads, lines with rounded tips or other end-marks ( Fig 6A ), and lines without any marks [ 17 ]. Remarkably, a single arrow type may convey distinct semantic meanings: an upward arrow may signify an upward movement, an increase, or a positive connotation, while a circular arrow can symbolize various temporal scales, from a day, to year, or an entire life cycle [ 17 ]. In many academic domains, arrows have also specialized applications, such as a corner/bent arrow that in molecular genetics illustrates transcription start sites [ 16 ]. Arrows can also depict various movements, representing phenomena like the passage of a molecule through a membrane, the migration of cells within a tissue, or the collective herd movement of animals. And finally, arrows and lines are also commonly used for labeling and directing attention to specific structures or regions of interest.

Common arrow types (A) and arrows in context (B).

https://doi.org/10.1371/journal.pcbi.1011789.g006

It is crucial that you clearly communicate the purpose of your arrows to your audience. When combining 2 different arrow types in a single graphical abstract, you should ensure they are visually distinct and explained. Moreover, the context in which an arrow is presented has substantial influence on how it’s perceived ( Fig 6B ).

Even with a clear layout and arrows, graphical abstracts can appear overwhelming. This feeling is rooted in the limitations of our visual system. Miller postulated the “Magical number 7,” suggesting that human sensory perception can effectively process only about 7 elements (plus or minus 2) at a time [ 18 ]. Of course, graphical abstracts typically comprise more than 7 elements. To address this challenge, design principles, often referred to as “Gestalt principles,” come into play, aiding in the organization of elements into interconnected units, or “chunks,” which enhances the information conveyed and reduces cognitive load [ 19 ].

Some of the design principles are especially helpful for graphical abstract design. “Proximity” suggests that elements can be grouped by minimizing their physical distance on the page. “Similarity” describes that elements form groups when they share common visual attributes. Such visual attributes, e.g., a shared color, pattern, or shape [ 20 ], may even lead to grouping when elements are not in close physical proximity. Grouping by similar appearance is helpful, e.g., in scatterplots, but can cause confusion if applied erroneously to non-grouped elements (see #8). “Closure” stipulates that elements within the same boundary are grouped, which explains the frequent use of boxes in design templates. However, it’s worth noting that boxes can often be replaced with white space to achieve a similar effect. The principle of “continuity” asserts that elements arranged along an invisible axis visually form a group, an idea that inspired Tufte to experiment with omitting x-axes in bar plots altogether [ 21 ]. And last, “similarity” suggests that elements arranged symmetrically appear grouped. These design principles are helpful for graphical abstracts but also valuable for improving your further designs such as scientific figures, as exemplified by Bang Wong [ 22 , 23 ].

Rule 7: Text

The most effective way to ensure audiences comprehend complex insights with graphical abstracts is by seamlessly integrating text and visuals [ 24 , 25 ]. To captivate your audience, consider incorporating well-known keywords and phrases [ 7 ]. Text can also serve as a substitution when suitable images or pictograms are unavailable, particularly for specialized names or terminology, e.g., “acetylcholine.” Text is also important for labeling ambiguous or unusual visuals, icons, or arrows. For example, a circle you use could represent a molecule, an area, or a cell. While text offers additional clarifications, be sure to keep your titles and annotations concise, devoid of jargon, and limited to common abbreviations, all of which in general enhance readability and citations of scientific articles in general [ 26 ]. Lastly, text can play a dual role as a legend when the annotation mirrors the encoding style of associated visual elements. You may color a key word in the title in the same hue as the associated data in the abstract (see Fig 3 ).

Rule 8: Colors

A key function of appealing colors in graphical abstracts is to engage your audience. Beyond that colors have further roles, color highlights, contrasts, encodes quantities, or represents the natural appearance of the depicted objects ( Fig 7A–7C ). When colors encode quantitative information, sequential or continuous data should be encoded with varying saturations of a single color, diverging data with e.g., two-color scheme, and for qualitative data you may vary the hue [ 27 , 28 ].

Color can highlight (A), encode numbers (B), or show natural appearance (C) in graphical abstracts. Be careful with your color choice when using a colored background. Image: Albrecht Dürer, Public domain, via Wikimedia Commons ( https://commons.wikimedia.org/wiki/File:Albrecht_D%C3%BCrer_-_Hare,_1502_-_Google_Art_Project.jpg ).

https://doi.org/10.1371/journal.pcbi.1011789.g007

Several tools are available that may be helpful when selecting your color schemes. Colorbrewer by Cynthia Brewer ( https://colorbrewer2.org/ ) is useful for choosing colors to encode numerical data, while Paletton ( https://paletton.com/ ) enables the selection of attractive color combinations using a color wheel. These tools can assist in achieving harmonious appearances through adjacent colors or creating striking contrasts by employing complementary colors.

Consistency in color usage is important (see #6, principle of similarity). It is vital that you maintain the same color code and scheme within the abstract, and between the abstract and the main manuscript. A change in color is not merely a shift in aesthetics, it signifies a change in meaning. Colors, being instantly perceptible, should be used sparingly to prevent overwhelming the audiences and diverting their attention from the primary message. Hence, make your color choices with utmost care.

When selecting colors, you should ensure that they are accessible to your color blind audiences [ 29 ]. But more generally, you should consider possible limitations to visually impaired audiences. A comprehensive study provides an overview of accessibility in visualizations for different target groups (i.e., color-blind, visually impaired, and blind individuals) and various visual tasks [ 30 ]. A few steps help to improve accessibility: all figures, including graphical abstracts, must always be described with accompanying text. You may also be able to provide Alt-text descriptions for screen reader software. Additionally, also visually able audiences differ in their perception of color and contrast and therefore color should be avoided as the sole channel for key information (see also #7, labeling visuals). Beyond avoiding certain color combinations, like red-green for individuals with Deuteranopia, also low-contrast color combinations and many background colors may reduce visibility and thus accessibility. You can use numerous web-based tools (e.g., https://www.color-blindness.com/coblis-color-blindness-simulator/ ) or render your monitor display settings to assess legibility. WebAIM suggests a minimum contrast ratio of 4.5 to 1 for foreground and background colors and provides a tool for assessing color combinations ( https://webaim.org/resources/contrastchecker/ ). Finally, maintaining a sufficiently high resolution is vital for ensuring accessibility, allowing your audiences to print or zoom in to your visualizations as needed.

Rule 9: Tools for graphical abstracts

Graphical abstracts are typically prepared with the same software as posters and figures. Suitable are commercial (e.g., Adobe Illustrator, CorelDraw, Affinity Designer) or open-source (e.g., Inkscape) vector-design software. Vector-based graphics programs are particularly useful as they allow for zooming in and out of visualizations without quality loss. For most graphical abstracts PowerPoint will also produce sufficient results, especially when the canvas size is adjusted and slides are exported as vector graphics such as PDF. When saving your graphical abstract make sure that your images are not compressed to prevent pixelation artifacts.

A comprehensive article reviews many common software used for illustrations as well as their advantages, disadvantages and pricing is available [ 31 ]. If you wish to use the free vector graphic software Inkscape, you may consult a practical guide for biologists [ 32 ]. Inkscape is rapidly developing and now allows direct import of icons from icon libraries, as well as processing of images and data with scripts inside the software. The proprietary alternative to Inkscape is Adobe Illustrator, which is widely adopted by scientists and for which tutorials are available [ 33 ]. Another commercial software is CorelDraw which can, like Inkscape, incorporate icons from many web-based icon libraries.

In recent years, several web-based drawing softwares have become available, such as Canva or Figma. BioRender is a proprietary web-based software powered by a large biomedical icon library, which is an attractive feature to its users; however, their appearance, shape, color, and detail cannot be changed. A drawback to many labs is also BioRender’s continuous adaptation of licenses, while an advantage is its interface with public databases, such as the Protein Data Bank. Another web-based tool is Mindthegraph, which also offers in addition design consulting. A summary of tools is available [ 7 ].

Pictograms and icons can be imported in all programs, including the web-based tools, as SVG or PNG (see #2) and Inkscape even allows the direct, web-based import from icon libraries such as Bioicons or Reactome.

Rule 10: Before, during, after: Feedback

Visual design is a dynamic and iterative process. Consequently, graphical abstracts should undergo several rounds of assessment and adjustment to avoid common pitfalls such as unclear reading directions [ 2 ] and inconsistencies in elements and style within the visualization.

Feedback can be actively sought and integrated at various stages: during the formulating of your key message, the drafting of your prototype, or the final polishing phase. As a best practice, the book Storytelling With Data in fact recommends allocating dedicated time for discussing the visualizations in every meeting [ 34 ]. As in every design of a human–computer interaction, also for graphical abstracts you may seek expert feedback, e.g., from a scientists or designers that regularly prepare graphical abstracts, as well as user feedback, e.g., from scientists or students who may read your paper.

General feedback principles [ 35 ] also apply to visual work. This means that feedback should be specific, tangible, and task-oriented and those seeking feedback should be clear in their request. In graphical abstracts, the audience must decode the visual representations. You can get feedback by observing how an expert or user is interacting with your graphical abstract, or by asking for their opinions. Ask what they see at first glance to see if the visual weight aligns with the key message. Ask about clarity of the layout and reading direction, including the meaning of arrows, and the comprehensibility of visual elements and colors. Alli Torban from Tableau, a visual design company, imparts additional guidance on the intricacies of soliciting and receiving feedback for visual designs [ 36 ].

When designing graphical abstracts in a team, we usually exchange rapid drafts or sketches of the graphical abstract several times before a solid idea emerges ( Fig 1 ) and is then prepared for publication [ 37 ]. In our experience, the process of preparing a graphical abstract also serves as a valuable exercise to assess whether our key message is succinct. It also aids writing teams and grant writers in aligning toward a shared vision or objective. The graphical abstract thus serves as a valuable tool for bridging communication or knowledge gaps in transdisciplinary teams such as consortia of clinicians, engineers, and biologists.

While initially graphical abstracts may seem like extra work for little reward, we hope that our 10 rules encourage you to start creating understandable and gorgeous graphical abstracts. A useful resource for educators wishing to teach graphical abstract preparation in a classroom setting is available from Agrawal and Ulrich, who provide templates for exercises and downloadable sample materials [ 9 ]. A quick guide, along with a PowerPoint template, is also available from Elsevier [ 38 ]. And for inspiration the British Medical Journal hosts a collection of infographics ( https://www.bmj.com/infographics ). Once you become familiar with the format of graphical abstracts, you may also wish to experiment with styles and forms. Usually, journals do not limit their authors: we have seen artistic, comic-style [ 39 ], and even hand-drawn (Fabio di Belvis: https://www.sciencedirect.com/science/article/pii/S0378517319307975?via%3Dihub ) graphical abstracts.

Acknowledgments

HKJ would like to acknowledge James P. Saenz for feedback on the draft version.

• 1. Lane S, Karatsolis A, Bui L. Graphical abstracts: a taxonomy and critique of an emerging genre. In: Proceedings of the 33rd Annual International Conference on the Design of Communication [Internet]. New York, NY, USA: Association for Computing Machinery; 2015 [cited 2023 Jun 4]. p. 1–9. (SIGDOC ‘15). https://doi.org/10.1145/2775441.2775465
• 2. Hullman J, Bach B. Picturing Science: Design Patterns in Graphical Abstracts. In: Chapman P, Stapleton G, Moktefi A, Perez-Kriz S, Bellucci F, editors. Diagrammatic Representation and Inference. Cham: Springer International Publishing; 2018. p. 183–200. (Lecture Notes in Computer Science).
• View Article
• PubMed/NCBI
• Google Scholar
• 4. Cell Press Graphical Abstract Guidelines. Cell Press [Internet]. Available from: https://www.cell.com/pb/assets/raw/shared/figureguidelines/GA_guide-1537202744020.pdf .
• 21. Tufte ER. The Visual Display of Quantitative Information [Internet]. Cheshire, Connecticut: Graphics Press; 2011 [cited 2023 Oct 23]. Available from: https://www.edwardtufte.com/tufte/books_vdqi?msclkid=d049a142816d107459dec4741f4dfaa5 .
• 34. Nussbaumer Knaflic C. Storytelling with Data: A Data Visualization Guide for Business Professionals. Wiley; 2015.
• 38. Graphical Abstract Template [Internet]. Available from: https://www.elsevier.com/journals/cellular-and-molecular-gastroenterology-and-hepatology/2352-345x/graphical-abstracts .

How to Create a Visual Abstract in 5 Easy Steps

Visual abstracts are becoming more and more popular ways to share research through social media. Using images can create a higher impact on platforms such as Twitter, LinkedIn, Instagram and Facebook. It can catch the eye of stakeholders that may not be used to reading through long research articles. Here is an example of a visual abstract we’ve designed and the steps we used to create it.

Working Sandwich Generation Women Utilize Strategies within and between Roles to Achieve Role Balance by Evans et. al. (2016) Visual Abstract Sample.

Create an outline of your research using as few words as possible.

This will help eliminate unnecessary text on a visual abstract that can be hard to read. There is a very short window to catch your audience’s attention and the less text the better.

2. Find the main takeaways or points your audience would want to know.

Think about your audience and what information they would want to understand and use in their clinical practice. They may not care what type of study design you used but specific clinical interventions and assessments are usually very interesting to practicing clinicians.

3. Use a templates to create your design.

Some examples of helpful resources include

Infographia

4. Find icons to replace chunks of words that represent the essence of those words.

You can find royalty free icons on:

The Noun Project

Creative Market (affiliate link)

5. Post where your audience spends their time.

If you’re looking to reach parents, a facebook groups for Mom’s might be the best place to share your work. If you want to reach more healthcare professionals looks into sharing at national conferences, if you want to reach out to students, Instagram is a more popular platform.

Still feeling stuck creating your design?

Use our free Visual Abstract Template.

We’d also love to help design and create it for you from start to finish. Connect with us below to learn more.

Who are stakeholders and why are they important.

Get in touch

555-555-5555

[email protected]

Limited time offer: 20% off all templates ➞

Best Graphical Abstract Examples with Free Templates

Creating engaging graphical abstracts can improve scientific publication rates and allows you to easily share your research in presentations and social media..

Graphical abstracts are becoming increasingly essential science communication tools for presentations and publications. Many academic journals now require scientists to submit graphical abstracts and the rise of social media has made summary images a standard expectation for communicating complex information. This article shows well-designed graphical abstract examples and also provides links to free templates.

What is a Graphical Abstract?

A graphical abstract is a visual representation of a research project. The goal of the abstract is to create a clear story of your scientific method and results that is quickly understood by your audience. The best graphical abstracts use a combination of data, illustrations, and formatting to make it easy to follow the main points of the research. Below is an example of a well-designed graphical abstract that uses left-to-right formatting to show the gathering initial data from TBI patients, treating patients with two treatment paths, and patient outcomes.

Learn how to design good graphical abstracts using PowerPoint and Adobe Illustrator

Graphical Abstract Examples

One great way to start brainstorming for your own graphical abstract is to look at examples to see which ideas and formats might best fit your own research story. Below is a review of the best graphical abstract examples, as well as links to download these free templates for your own designs.

Left-to-Right Designs

My top recommended graphical abstract design uses bold title text with left-to-right formatting for the details below it. This format is easy for people to understand and can be used to compare methods to results, describe a sequence of events, or show a series of scientific conclusions. Below are examples of my recommended left-to-right designs with 1-4 columns.

Click here to download these free graphical abstract templates for Adobe Illustrator and PowerPoint

Top-to-Bottom Designs

Another good option is to use a top-to-bottom formatting. This is an especially good design idea if your data output goes from a large quantity to a small quantity or if the research results naturally go from top to bottom, such as north to south on a map or from the atmosphere to the Earth. Below are examples of top-to-bottom graphical abstract designs with 1-4 rows.

Circular and Unique Graphical Abstracts

The final recommended formats are circular and unique formats such as timelines and Venn diagrams. These are less commonly used and should only be selected if the summary of your research is easier to understand using one of these designs than the left-to-right formatting.

Design Tools to Customize Graphical Abstracts

Knowing how to use design tools to create custom graphical abstracts has become an increasingly essential skill for researchers. Below is an example of a graphical abstract design that was customized using biological diagram templates and a list of the top design tools that scientists use to create graphical abstracts and scientific illustrations.

Adobe Illustrator

• Top recommended software for advanced scientific and graphic design. This is the digital design tool used by most professional scientific illustrators.
• This tool allows for full customization of graphical abstracts by creating high resolution vector designs where every pixel can be adjusted to make the perfect final design.
• Learn more about how to get Adobe Illustrator as a student or scientist .
• Costs: $240-252 for annual subscription

Affinity Designer

• Design software that is similar to Adobe Illustrator but with slightly fewer design features. This is a good affordable alternative to Adobe Illustrator.
• This tool allows for customization of graphical abstracts by creating high resolution vector designs where every pixel can be adjusted to make the perfect final design.
• Visit here to purchase the software: https://affinity.serif.com/en-us/designer/
• Cost: $70 one time payment

• PowerPoint is a commonly used software for scientists and has become increasingly good at allowing researchers to make custom designs using their shapes, lines and arrow features.
• This tool has limited design features, but these are not always needed if you know how to use PowerPoint well.
• Visit this page to learn more about purchase options.
• Cost: Free versions and $70-160 for full software

• Google Slides and Google Drawing are comparable tools to Microsoft PowerPoint. Scientists do not use these as often as PowerPoint, but it is still a good software to use if you are more familiar with Google products. 
• The design features are limited compared to Adobe Illustrator and Affinity Designer, but you can still use this software to create high quality graphical abstracts.
• Cost: Free with Google account

There are also tools such as BioRender that allow you to create graphical abstracts with images that you can copy/paste into designs. However, this tool has limited customization options and is very expensive if you want to download your work as high resolution images that are used for publications and presentations. Read this article to learn more about the costs, pros, and cons of popular scientific design tools . 

Use Graphical Abstracts to Promote Research

There are many different options to share your research with the public and your peers. Having a well-designed graphical abstract makes it easy to format the designs to share via presentations, scientific websites, and social media. This is a great way to increase interest in and awareness of scientific research.

In order to share your graphical abstract via social media, you may need to adjust your designs so that the image can be best formatted for different platforms. Each social media platform has their own preferred dimensions for the images you share. For example, if you want to share your graphical abstract on both Instagram and LinkedIn, you will want to adjust one version to fit a square image for Instagram and you probably won't need many adjustments to share a landscape image on LinkedIn. Below are examples of graphical abstract image formatting for social media posts on Instagram, LinkedIn, Facebook, and Twitter. 

Graphical Abstract Design Summary

All of the examples and tools described in this article can help you design impressive graphical abstracts and share them with a wider audience. Use the simple process below to start your own design.

• Step 1. Choose a design plan that looks good to you, best represents your data, and matches your intended scientific journal's formatting requirements. 
• Step 2. Create a draft of your design by drawing on paper or use digital design tools such as Adobe Illustrator, Affinity Designer, or PowerPoint to arrange your illustrations, text, and graphs. Learn more about graphical abstract design options by clicking on the resources below:
• Download free graphical abstract templates and view other science images  
• Sign up for free courses on graphical abstract and scientific illustration
• Step 3. Adjust the design formatting and colors until the main story of your research is clear.
• Read articles to learn more about data visualization design best practices and data storytelling
• Step 4. Share with scientists and the public via presentations, scientific websites, and social media. 

Create professional science figures with illustration services or use the online courses and templates to quickly learn how to make your own designs.

Interested in free design templates and training.

Explore scientific illustration templates and courses by creating a Simplified Science Publishing Log In. Whether you are new to data visualization design or have some experience, these resources will improve your ability to use both basic and advanced design tools.

Interested in reading more articles on scientific design? Learn more below:

Scientific Presentation Guide: How to Create an Engaging Research Talk

Data Storytelling Techniques: How to Tell a Great Data Story in 4 Steps

Best Science PowerPoint Templates and Slide Design Examples

Content is protected by Copyright license. Website visitors are welcome to share images and articles, however they must include the Simplified Science Publishing URL source link when shared. Thank you!

Online Courses

Stay up-to-date for new simplified science courses, subscribe to our newsletter.

Thank you for signing up!

You have been added to the emailing list and will only recieve updates when there are new courses or templates added to the website.

We use cookies on this site to enhance your user experience and we do not sell data. By using this website, you are giving your consent for us to set cookies: View Privacy Policy

Simplified Science Publishing, LLC

• Search Menu
• Sign in through your institution
• Advance articles
• Editor's Choice
• Supplement Archive
• Cover Archive
• IDSA Guidelines
• IDSA Journals
• The Journal of Infectious Diseases
• Open Forum Infectious Diseases
• Photo Quizzes
• State-of-the-Art Reviews
• Voices of ID
• Author Guidelines
• Open Access
• Why Publish
• IDSA Journals Calls for Papers
• Advertising and Corporate Services
• Advertising
• Journals Career Network
• Reprints and ePrints
• Sponsored Supplements
• Branded Books
• About Clinical Infectious Diseases
• About the Infectious Diseases Society of America
• About the HIV Medicine Association
• IDSA COI Policy
• Editorial Board
• Self-Archiving Policy
• For Reviewers
• For Press Offices
• Journals on Oxford Academic
• Books on Oxford Academic

Article Contents

Theory supporting visual media, utility of infographics, creating effective infographics, conclusions.

• < Previous

Creating Effective Infographics and Visual Abstracts to Disseminate Research and Facilitate Medical Education on Social Media

• Article contents
• Figures & tables
• Supplementary Data

Jennifer O Spicer, Caroline G Coleman, Creating Effective Infographics and Visual Abstracts to Disseminate Research and Facilitate Medical Education on Social Media, Clinical Infectious Diseases , Volume 74, Issue Supplement_3, 15 May 2022, Pages e14–e22, https://doi.org/10.1093/cid/ciac058

• Permissions Icon Permissions

Presenting information in a visual format helps viewers digest complex concepts in an efficient, effective manner. Recently, infographics have been used on social media and other digital platforms to educate health professionals, trainees, and patients about medical and public health topics. In addition, visual abstracts, visual representations of a research article’s written abstract, have been increasingly used to disseminate new research findings to other health professionals. In this review article, we will define infographics and visual abstracts, describe why they are useful, outline how to create them, and explain how researchers, educators, and clinicians can use them effectively. We share resources and a stepwise approach that allows readers to create their own infographics and visual abstracts for research dissemination, medical education, and patient communication.

With the advent of social media, using visual methods of communication has become increasingly important. Many scientists and health professionals feel comfortable preparing figures, tables, and slideshow presentations but may be less familiar with creating newer visual media, such as infographics and visual abstracts. Learning how to develop these visual tools will enhance their ability to share their scientific work and educate other health professionals and the general public.

Infographics are visual representations of information using a combination of charts, icons, or illustrations with minimal text. Visual abstracts, also known as graphical abstracts, are similar to infographics, but this term specifically refers to concise visual summaries of the main findings of an article. They were first popularized by Andrew Ibrahim [ 1 ] and have now been adopted by many journals [ 2 ]. Although some journals have staff who create visual abstracts, most require authors to submit their own. Visual abstracts are frequently posted on social media to highlight an article’s main findings. In this article, we will use the term “infographic” as an all-encompassing term referring to any visual used to represent information, including visual abstracts. We will use the term “visual abstracts” only when discussing content specific to visual abstracts. In this review, we will discuss the utility of infographics, outline principles for creating them, and provide resources for those interested in developing this skill.

Visuals help us interpret complex information more quickly than text alone [ 3 , 4 ]. Intuitively, most of us understand this based on our experiences watching slideshow presentations. Simple slides with relevant graphics and limited text make it easier to digest complex information. Presenting information in a visual manner decreases the cognitive load, or “mental energy,” required to interpret information [ 5 ]. Cognitive load theory describes 3 components that contribute to this mental energy: (1) intrinsic load, which refers to the inherent complexity of a topic; (2) extraneous load, which refers to external factors that affect learning (eg, distracting images not relevant to the topic); and (3) germane load, which refers to the mental energy expended to organize and understand content [ 6 ]. The cognitive theory of multimedia learning explains why adhering to certain multimedia principles reduces cognitive load and enhances learning [ 5 , 7 ].

Well-crafted visuals reduce intrinsic load, optimize germane load, and eliminate extraneous load. Intrinsic load is reduced when using visuals because images allow our brains to store information differently, as explained by dual coding theory [ 8 ]. Dual coding theory describes how our brain processes information using 2 channels—a verbal channel for processing language and a nonverbal channel for processing other stimuli, such as images or sounds [ 9 ]. When we see a picture of something, our brain stores that information as a visual image but also as a word in our language center, thus making it easier to remember and retrieve information in the future. Therefore, including images is more effective than just using text, as highlighted by data visualization experts, such as Edward Tufte, Stephanie Evergreen, and David McCandless; however, it is important for visuals to include only relevant information to avoid adding extraneous load. Carefully organizing materials within visual materials facilitates comprehension, thus optimizing germane load. For example, consider John Snow’s famous epidemiological map of the London cholera outbreak ( https://www.theguardian.com/news/datablog/2013/mar/15/john-snow-cholera-map ) or Florence Nightingale’s polar area diagram of causes of death ( https://www.davidrumsey.com/luna/servlet/detail/RUMSEY~8~1~327826~90096398:Diagram-of-the-Causes-of-Mortality-;JSESSIONID=1d4a4d62-1b43-4dd3-8cb6-d77e1ee66910 ) among the British Army. Figure 1 demonstrates multimedia principles relevant to infographics [ 5 , 7 ].

Explanation and illustration of multimedia design principles that can decrease cognitive load. Abbreviations: HIV, human immunodeficiency virus; INSTIs, integrase strand transfer inhibitors; NNRTIs, nonnucleoside reverse-transcriptase inhibitors; NRTIs, nucleoside reverse-transcriptase inhibitors. Image created with BioRender.com.

Infographics can be used for a variety of purposes. Educators, clinicians, and public health professionals frequently use them to communicate established knowledge to learners, patients, and the general public. Researchers, in contrast, may benefit more from creating visual abstracts to share and disseminate new research findings.

Educators, Clinicians, and Public Health Professionals

Adhering to research-based principles for multimedia learning when developing slideshow presentations has been shown to result in superior knowledge retention [ 10 ]. These same principles can be applied to the development of infographics to facilitate understanding of new content. Medical educators can use infographics to provide a succinct overview of a topic, making it easier for learners to learn and retain information. Creating a summary image for a topic helps learners progressively add new details to an underlying mental framework, thus scaffolding their learning. For examples, educators, such as the Clinical Problem Solvers ( https://clinicalproblemsolving.com/ ), have created visual summaries of information such as illness scripts and diagnostic schema that are widely shared over social media.

Learners, however, can also benefit from creating infographics themselves. For example, asking learners to create a visual abstract of a recent journal article to present at journal club helps them identify and summarize the important information, thus enhancing retention. When creating visual summaries of information, learners organize and build connections between content, and this elaboration on their knowledge facilitates information retention [ 11 ]. In addition, students learn how to communicate scientific information more effectively [ 12 ].

Clinicians and public health professionals can also use infographics to communicate with patients. Using infographics for patient education has been shown to improve health knowledge and outcomes [ 13 , 14 ]. Public health organizations—such as the Centers for Disease Control and Prevention (CDC) ( https://www.cdc.gov/socialmedia/tools/InfoGraphics.html ), the World Health Organization ( https://www.who.int/multi-media ), and the American Public Health Association ( https://www.apha.org/news-and-media/multimedia/infographics )—develop and share infographics to educate the general public. Infographics have the added benefit of being easily understood regardless of primary language or level of education.

Researchers

Research impact has traditionally been defined by the “prestige” of the journal in which an article is published (ie, the impact factor) and the number of citations that an article receives. More recently, however, altmetrics (short for “alternative metrics”) have been recognized as a method to measure how widely an article is disseminated over social media, which increases its visibility to other health professionals and the general public [ 15 ].

Using images increases the impact of content shared on social media. Most studies focus on Twitter, and they have shown that adding images to posts increases altmetrics, including the number of people who view a post (ie, “impressions”), interact with a post by “liking” it or clicking on it (ie, “engagement”), or share a post (ie, “retweet”) [ 16 , 17 ]. This increased visibility increases the number of people who click on the article link [ 16 , 18 ], although this does not always translate to increased full-text views of the article [ 19 ]. Using visual abstracts in tweets, has been shown to result in more interaction than just using a key figure from the article [ 16 , 17 ]. Thus, using visual abstracts and disseminating them on social media is one way to increase the attention that a research article receives, especially if the post is on an account with a large number of followers.

One concern that has been raised with visual abstracts is the tendency for them to oversimplify an article [ 2 , 20 ]. This is a valid concern; however, it is similar to what already occurs with a written abstract. The visual abstract is meant to serve as a preview on an article, similar to a written abstract, not as a replacement for reading the full article [ 1 ]. To avoid misrepresenting articles, it has been suggested that visual abstracts should be part of the peer review process or approved by the author or journal as a fair representation of the data [ 2 , 20 ].

Developing effective infographics requires combining principles from the fields of scientific communication and graphic design. The goal is to create visual material that communicates information accurately yet facilitates rapid understanding of key points. The sections below outline how to build infographics, as summarized in Figure 2 . These recommendations are based on a compilation of research-based findings from graphic design and scientific communication literature and recommendations from others with experience in building infographics [ 1 , 2 , 21–23 ]. Those readers familiar with design thinking will recognize the similarities between those principles—empathize, define, ideate, prototype, and test—with the process we recommend here [ 24 ].

Illustration of the steps used to create an infographic or visual abstract. Note that although this figure is also used as the visual abstract for the current article, this practice is not typically recommended.

Outline the Content

The first step in creating an infographic is defining the primary goal, the main message, and the target audience. The primary goal refers to the conceptual rationale for creating the infographic. Is it intended to summarize a topic or article, explain an abstract concept, compare and contrast 2 entities, outline a process, describe changes over time, contextualize a statistic, and/or persuade viewers to perform a certain action? The underlying purpose determines what supporting content should be included and how it should be organized.

Next, summarize the key message in a single sentence. What is the one “take home” point someone should understand after looking at it? All keywords should be included in this message, just as the title of a manuscript should include all key search terms. This sentence can later be used as a social media post or summary sentence within the infographic. This step is critical before developing the visual design, because it is easy to be distracted from the primary message of the infographic during the design process.

Finally, consider the target audience for the infographic. The content of the infographic, including the language and depth of content, will differ depending on whether the primary audience is a layperson, a learner, or an expert in a field. For those creating materials for an audience without a scientific background, the CDC’s Health Communication Playbook provides helpful guidelines to improve communication [ 25 ]. The CDC Clear Communication Index ( https://www.cdc.gov/ccindex/widget.html ) is a tool containing 24 questions based on scientific communication literature that can be used to evaluate health communication materials and determine whether they need to be revised to enhance understanding [ 26 ]. The CDC also has a tool on their Web site that translates medical jargon into more accessible language ( https://www.cdc.gov/healthcommunication/everydaywords/ ).

After determining the primary goal, main message, and audience, write a short list of supporting information that needs to be included in the infographic. Include Only critical details needed to support the main point. Brevity is key, especially if the infographic is being shared on social media.

Sketch the Layout

Once the content has been outlined, it is time to develop an initial layout for the infographic. Graphic design is a complex field with a wealth of evidence-based research supporting principles of effective design. Most health professionals do not have formal training in graphic design principles, but exploring some readable books on the topic can provide tips [ 27–29 ].

An organized layout is a key component of an effective infographic. The structure of the layout should match the primary goal of the infographic, as demonstrated in Figure 3 . Use shapes, such as boxes, to group content and structure material within the layout. Orient the boxes along a grid, using horizontal and vertical lines to establish sections, to align the content and promote organization and cohesion [ 28 ]. Leave unfilled space—termed “white space” or “negative space”—between objects to avoid clutter and allow viewers to focus on the primary content; however, ensure that spacing is consistent. Finally, establish a visual hierarchy to ensure that viewers know how to navigate through the infographic. Viewers tend to scan content similarly to how they would read narrative prose, although their viewing patterns are affected by item complexity. Size, shape, color, and images can all affect viewers’ interpretation of the visual hierarchy. Add elements such as numbers, arrows, and headings to guide viewers on the order in which they should read content.

Examples of potential layouts for infographics based on their underlying primary goal. The image in example 1 was created by Emma Levine, MD, for the Clinical Problem Solvers ( https://clinicalproblemsolving.com/ ) and has been reused with their permission. The image in example 2 was created by Miriam Ahmed, PharmD, and has been reused with her permission.

Looking at examples of other infographics can provide some initial ideas on a general layout. Infographics and visual abstracts can be found by searching these terms within a Web-based search engine or by searching the hashtags #infographic, #VisualAbstract, or #GraphicalAbstract on social media sites. GrepMed ( https://www.grepmed.com/ ) is a searchable collection of community-sourced medical images and infographics. The Febrile podcast has created summary infographics on infectious diseases topics discussed on their podcast ( http://febrilepodcast.com/infographics/ ), and the Clinical Problem Solvers create infographics of their diagnostic schemas and illness scripts on their Web site, including some topics in infectious diseases ( https://clinicalproblemsolving.com/ ). Many medical education sites, public health organizations, podcasts, and journals host collections of infographics or visual abstracts on their Web sites. The Visual Abstract Database ( https://visualabstract.pro/ ) contains a gallery of visual abstracts from a variety of journals. Figure 4 shows an example layout for a visual abstract.

Sample layout for a visual abstract.

When developing the layout, resist the temptation to open slideshow software on a computer. Instead, use a pencil and a blank sheet of paper to create rough sketches. Drawing frees the user from default parameters present in slideshow software, thus promoting creativity and preventing frustration that may occur when trying to determine how to accomplish specific goals within a software program. The drawing does not have to be perfect. It just needs to outline a general idea for the design.

Choose a Platform

After sketching an initial draft on paper, transfer it to a digital format to refine it. Many software platforms exist that include premade infographic templates and graphics that can be included in the illustrations ( Table 1 ). A few of the platforms have a free version, although many require a subscription to be able to download and publish infographics. Microsoft Powerpoint, Keynote, and GoogleSlides are commonly used and have all of the tools necessary to create an infographic.

Examples of Software Programs Available for Developing Infographics a

The examples in this table are not an exhaustive list of options.

$ indicates less than $200 per year and $$ indicates more than $200 per year.

The number of $ signs correlates with the cost of the product.

Select a Color Scheme

Color choice is often a personal preference. Although some colors tend to evoke specific emotions or meanings, those interpretations vary based on an individual’s culture, personal history, and preferences [ 30 , 31 ]. David McCandless has created an infographic highlighting how colors are interpreted differently based on culture [ 32 ], which can influence implicit interpretations of visualizations. When choosing a color combination, also consider accessibility, as approximately 10% of the population has some type of color vision deficit, most commonly the inability to distinguish between red and green [ 33 ]. It is generally recommended to use 3–5 colors, including at least 1 light, 1 dark, and 1 emphasis color [ 28 , 29 ]. Many Web sites exist that can aid in the selection of a cohesive color combination using principles of color theory. For example, the Adobe Color site ( https://color.adobe.com ) will create a color palette from an image or a base color, and the color codes can then be used in another software program; moreover, it can even simulate how color palettes will appear to individuals with common color vision deficiencies.

Color use within infographics should be purposeful and planned. Ensuring color contrast is essential. A light background with dark text and graphics is often easier to read than light text on a dark background; color text on a color background is more difficult to read and should be avoided [ 29 , 34 ]. Graphic designers often advocate the 60-30-10 rule of color use, which advocates for using a dominant (often neutral) color for 60% of the space, a secondary color for 30%, and an accent color for 10% [ 35 ]. Any additional color use should be limited and applied meaningfully to help individuals better interpret the data presented [ 29 ].

Incorporate Graphics

After sketching the layout, consider how to present content visually, prioritizing graphics over words. Any image included should serve a purpose and not be merely decorative. Including irrelevant graphics distracts learners from the primary message and hinders learning [ 5 , 7 ]. In general, it is best to use the simplest graphic possible while including any necessary details, as filtering out any extraneous elements improves comprehension [ 28 , 36 ]. Any text included should be brief, using minimal words while maintaining clarity.

Examples of graphics include icons, medical illustrations, timelines, tables, charts, and graphs. Each type serves a different purpose, and a single infographic often includes multiple graphic elements ( Figure 5 ). Of note: if tables, graphs, or charts are used, they should be much simpler than those presented in a scientific manuscript and should only contain critical results.

Examples of how graphic elements can be incorporated into infographics.

Icons and medical illustrations are commonly used in infographics. Icons are simplified pictorial representations of objects that may be in color or black and white. Icons are available for free in many of the software programs previously mentioned; however, other Web sites, such as the Noun Project ( https://thenounproject.com/ ) and FlatIcon ( https://www.flaticon.com/ ), have additional icons available free or through a subscription service, depending on how they will be used. When choosing an icon, ensure that the meaning is clear or relevant words are nearby. Cultural conventions may affect the interpretation of icons and result in different interpretations [ 37 ]. More complex medical illustrations are also available on other Web sites, including Servier Medical Art ( https://smart.servier.com/ ), a free service, and BioRender ( https://biorender.com/ ), which provides more options but requires a paid subscription. Choosing consistent images (ie, similar styles of icons or illustrations) in a single infographic improves cohesiveness and aesthetics.

Other graphic elements can be used to communicate more complex concepts. Flowcharts or numbered illustrations are helpful for explaining a process. Tables and bar graphs are helpful for making comparisons, such as outcomes across 2 groups in a study, and these are more easily interpreted than pie charts, which should generally be avoided [ 38 ]. Line graphs can be used to show trends in data. Two-dimensional graphics should be used instead of 3-dimensional ones, as the latter increase complexity and make it more difficult for individuals to make visual comparisons [ 38–40 ].

If using a chart, it is important to ensure that the chart type selected matches the primary goal of the visual (eg, comparing 2 entities or showing change over time). Selecting the best method for data presentation is a skill, and there are resources to help guide chart selection based on the intended goal [ 27 , 38 ]. Figures from a manuscript should not just be reused in an infographic. All extraneous information should be removed from the graph, including unnecessary numbers, words, and lines, in order to emphasize the main point. Simplicity is key.

Finally, check the copyright license before using an image. Most software programs allow use of their images and products on social media sites and in publications, although some require a subscription to download materials. Web sites containing collections of icons and illustrations have information on their sites regarding how images can be used. If searching for images using an internet search engine, check the license of the images before using them. Some images have Creative Commons copyright licenses that allow reuse with minimal restrictions.

Ask for Feedback

After creating an infographic, it is helpful to ask for feedback from others to confirm that the message is clear and concise. Having someone from the target audience review the infographic is essential to ensure that it is interpreted accurately, especially for infographics developed for patients or the general public [ 41 ]. Collaborating with a graphic artist can result in an even better product [ 42 ].

Infographics provide a powerful method for communicating complex information. Researchers, educators, clinicians, and public health professionals should consider how they can use infographics and visual abstracts to communicate information to their target audience. Sharing infographics on social media results in wider dissemination of materials, increasing the impact of health professionals beyond their local institution. When designing an infographic, consider research-based principles of scientific communication and graphic design.

Acknowledgments. The authors thank Varun Phadke, MD, who provided input on an early draft of the manuscript.

Supplement sponsorship. This supplement is supported by the Infectious Diseases Society of America.

Potential conflicts of interest. All authors: No reported conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

Ibrahim A , Abbass MA , Anderson P , et al.  . Use of a visual abstract to disseminate scientific research, version 4 . Available at: https://static1.squarespace.com/static/5854aaa044024321a353bb0d/t/5a527aa89140b76bbfb2028a/1515354827682/VisualAbstract_Primer_v4_1.pdf . Accessed 4 March 2021 .

Ramos E , Concepcion BP. Visual abstracts: redesigning the landscape of research dissemination. Semin Nephrol 2020 ; 40 : 291 – 7 .

Google Scholar

Larkin JH , Simon HA. Why a diagram is (sometimes) worth ten thousand words. Cogn Sci 1987 ; 11 : 65 – 100 .

Vekiri I. What is the value of graphical displays in learning? Educ Psychol Rev 2002 ; 14 : 261 – 312 .

Mayer RE. Applying the science of learning to medical education. Med Educ 2010 ; 44 : 543 – 9 .

van Merriënboer JJG , Sweller J. Cognitive load theory in health professional education: design principles and strategies. Med Educ 2010 ; 44 : 85 – 93 .

Mayer RE , Moreno R. Nine ways to reduce cognitive load in multimedia learning. Educ Psychol 2003 ; 38 : 43 – 52 .

Paivio A , Csapo K. Picture superiority in free recall: imagery or dual coding? Cognit Psychol 1973 ; 5 : 176 – 206 .

Clark JM , Paivio A. Dual coding theory and education. Educ Psychol Rev 1991 ; 3 : 149 – 210 .

Issa N , Schuller M , Santacaterina S , et al.  . Applying multimedia design principles enhances learning in medical education. Med Educ 2011 ; 45 : 818 – 26 .

Cutting MF , Saks NS. Twelve tips for utilizing principles of learning to support medical education. Med Teach 2012 ; 34 : 20 – 4 .

Shanks JD , Izumi B , C S , Martin A , Byker Shanks C. Teaching undergraduate students to visualize and communicate public health data with infographics. Front Public Health 2017 ; 5 : 315 .

Stonbraker S , Liu J , Sanabria G , et al.  . Clinician use of HIV-related infographics during clinic visits in the Dominican Republic is associated with lower viral load and other improvements in health outcomes. AIDS Behav 2021 ; 25 : 4061 – 73 .

Egan M , Acharya A , Sounderajah V , et al.  . Evaluating the effect of infographics on public recall, sentiment and willingness to use face masks during the COVID-19 pandemic: a randomised internet-based questionnaire study. BMC Public Health 2021 ; 21 : 367 .

Trueger NS , Thoma B , Hsu CH , Sullivan D , Peters L , M L. The altmetric score: a new measure for article-level dissemination and impact. Ann Emerg Med 2015 ; 66 : 549 – 53 .

Oska S , Lerma E , Topf J. A picture is worth a thousand views: a triple crossover trial of visual abstracts to examine their impact on research dissemination. J Med Internet Res 2020 ; 22 : e22327 .

Wadhwa V , Latimer E , Chatterjee K , McCarty J , Fitzgerald RT. Maximizing the tweet engagement rate in academia: analysis of the AJNR Twitter feed. AJNR Am J Neuroradiol 2017 ; 38 : 1866 – 8 .

Ibrahim AM , Lillemoe KD , Klingensmith ME , Dimick JB. Visual abstracts to disseminate research on social media: a prospective, case-control crossover study. Ann Surg 2017 ; 266 : e46 – 8 .

Huang S , Martin LJ , Yeh CH , et al.  . The effect of an infographic promotion on research dissemination and readership: a randomized controlled trial. CJEM 2018 ; 20 : 826 – 33 .

Ibrahim AM. Seeing is believing: using visual abstracts to disseminate scientific research. Off J Am Coll Gastroenterol ACG 2018 ; 113 : 459 – 61 .

Hernandez-Sanchez S , Moreno-Perez V , Garcia-Campos J , Marco-Lledó J , Navarrete-Muñoz EM , Lozano-Quijada C. Twelve tips to make successful medical infographics. Med Teach 2021 ; 43 : 1353 – 9 .

Aoki S. Design tips for graphical abstracts. BioRender Learning Hub, 2021 . Available at: https://learn.biorender.com/tutorial/designing-graphical-abstracts . Accessed 4 March 2021 .

Coleman C. How to create a visual abstract. YouTube. 2021 . Available at: https://www.youtube.com/watch?v=XNmCc2FehqE . Accessed 4 March 2021 .

Sandars J , Goh PS. Design thinking in medical education: the key features and practical application. J Med Educ Curric Dev 2020 ; 7 : 2382120520926518 .

Agency for Toxic Substances and Disease Registry, Centers for Disease Control and Prevention, US Department of Health and Human Services . Health Communication Playbook: resources to help you create effective materials . 2018 . Available at: https://www.cdc.gov/nceh/clearwriting/docs/health-comm-playbook-508.pdf . Accessed 4 March 2021 .

Baur C , Prue C. The CDC Clear Communication Index is a new evidence-based tool to prepare and review health information. Health Promot Pract 2014 ; 15 : 629 – 37 .

Tufte ER. The visual display of quantitative information . Cheshire, CT : Graphics Press , 2001 .

Google Preview

Malamed C. Visual design solutions: principles and creative inspiration for learning professionals . Hoboken, NJ : John Wiley & Sons , 2015 .

Evergreen S. Presenting data effectively: communicating your findings for maximum impact . 2nd ed. Thousand Oaks, CA : Sage Publishing , 2018 .

Clarke T , Costall A. The emotional connotations of color: a qualitative investigation. Color Res Appl 2008 ; 33 : 406 – 10 .

Taylor C , Clifford A , Franklin A. Color preferences are not universal. J Exp Psychol Gen 2013 ; 142 : 1015 – 27 .

McCandless D ; AlwaysWithHonor.com . Colours in culture . Available at: https://informationisbeautiful.net/visualizations/colours-in-cultures/ . Accessed 9 January 2022 .

Swanson WH , Cohen JM. Color vision. Ophthalmol Clin N Am 2003 ; 16 : 179 – 203 .

Richardson RT , Drexler TL , Delparte DM. Color and contrast in E-learning design: a review of the literature and recommendations for instructional designers and Web developers. MERLOT J Online Learn Teach 2014 ; 10 : 657 – 70 .

Adelugba A. How the 60-30-10 rule saved the day. UX Collect . 2020 . Available at: https://uxdesign.cc/how-the-60-30-10-rule-saved-the-day-934e1ee3fdd8 . Accessed 4 March 2021 .

Strobel B , Lindner MA , Saß S , Köller O. Task-irrelevant data impair processing of graph reading tasks: an eye tracking study. Learn Instr 2018 ; 55 : 139 – 47 .

Mangan J. Cultural conventions of pictorial representation: iconic literacy and education. Educ Commun Technol 1978 ; 26 : 245 – 67 .

Evergreen S. Effective data visualization: the right chart for the right data . 2nd ed. Thousand Oaks, CA : Sage Publishing , 2019 .

Larsen T. Lies, damned lies, statistics & computer graphics. YouTube. 2016 . Available at: https://www.youtube.com/watch?v=hsWs6d22dPY . Accessed 9 January 2022 .

Miliard M. Best practices for healthcare data visualization . Healthcare IT News . 2016 . Available at: https://www.healthcareitnews.com/news/best-practices-healthcare-data-visualization . Accessed 9 January 2022 .

Arcia A , Suero-Tejeda N , Bales ME , et al.  . Sometimes more is more: iterative participatory design of infographics for engagement of community members with varying levels of health literacy. J Am Med Inform Assoc JAMIA 2016 ; 23 : 174 – 83 .

Khoury CK , Kisel Y , Kantar M , et al.  . Science-graphic art partnerships to increase research impact. Commun Biol 2019 ; 2 : 295 .

• education, medical
• health personnel
• public health medicine
• social media

Email alerts

More on this topic, related articles in pubmed, citing articles via, looking for your next opportunity.

• Recommend to your Library

Affiliations

• Online ISSN 1537-6591
• Print ISSN 1058-4838
• Copyright © 2024 Infectious Diseases Society of America
• About Oxford Academic
• Publish journals with us
• University press partners
• What we publish
• New features  
• Open access
• Institutional account management
• Rights and permissions
• Get help with access
• Accessibility
• Media enquiries
• Oxford University Press
• Oxford Languages
• University of Oxford

Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide

• Copyright © 2024 Oxford University Press
• Cookie settings
• Cookie policy
• Privacy policy
• Legal notice

This Feature Is Available To Subscribers Only

Sign In or Create an Account

This PDF is available to Subscribers Only

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

advertisement

SYSTEM UPGRADE NOTICE

We have made some exciting digital upgrades! All members and customers will need to reset their passwords to access their accounts in our new system. Doing so will allow you to complete transactions and access all AUA websites, including UrologyHealth.org, The Journal of Urology and AUA University, as well as all mobile apps.

Reset your password now .

Attention: Restrictions on use of AUA, AUAER, and UCF content in third party applications, including artificial intelligence technologies, such as large language models and generative AI. You are prohibited from using or uploading content you accessed through this website into external applications, bots, software, or websites, including those using artificial intelligence technologies and infrastructure, including deep learning, machine learning and large language models and generative AI.

Creating visual abstracts to promote your research: A how-to guide

Denzel Zhu 1 , Alex Sankin, MD 1,2

• Albert Einstein College of Medicine, Bronx, NY
• Department of Urology, Montefiore Medical Center, Bronx, NY

Contact information: [email protected] , [email protected]

Social media (e.g. Twitter, Facebook, Instagram) are platforms heavily used by medical professionals, and the lay public, to communicate and share results from research studies. Visual abstracts, which are visual representations of research articles, are visually engaging and concise ways of sharing the results of studies on social media. The use of visual abstracts to promote research, as measured by impressions on Twitter, has been shown to dramatically increase the numbers of individuals who review and learn about the research. 1,2 However, there are few guidelines on creating effective visual abstracts. Our goal here is to provide guidance on creating an effective visual abstract, based on our past experience. 3

Abstract Design

We used Microsoft PowerPoint v16.64 (Microsoft, Redmond, WA)’s default slide size of 16:9 (Widescreen) to create our abstract. In general, we divide the slide into three major sections: the header (where the journal logo and visual abstract title go), the body of the abstract (where the study’s methodology, exposure, and outcomes are listed), and the footer (where the study’s conclusion, citation, and abstract author’s byline and contact information are listed, as well as institutional logos from the author’s institution; Figure 1).

Figure 1. Structure of the visual abstract.

For the title of the abstract, we recommend against replicating the study title for the visual abstract. Rather, we suggest writing the central question that the study is attempting to answer. For example, our recent study’s title was “Finasteride and risk of bladder cancer in a multiethnic population”. However, the central question our study was attempting to answer was: “Did the chemoprotective effect of finasteride for the prevention of bladder cancer vary by race/ethnicity?” The question provides more context for viewers of the visual abstract, and creates a narrative tension that the remainder of the visual abstract can provide the answer to.

For the body of the abstract, we suggest using the methodology space to describe overall characteristics of the study population, such as the number of participants, their demographics, and the inclusion criteria. For the exposure and outcome spaces, we suggest filling these spaces with the exposure of the study (e.g. drug vs placebo) and the relevant outcome of the study (e.g. the risk of developing bladder cancer). In the footer of the abstract, we suggest placing the major conclusion of the study, along with information for viewers to find more information about the study and the author’s institution and contact information (i.e., Twitter handle).

Style Choices

We suggest using large, bolded, easily readable fonts to write out information within the visual abstract. Given that many abstracts will be viewed on mobile devices, or compressed when uploaded to social media networks, there will be limits on how much text which can be included in the abstract. In the abstract below, we used size 22 font, Arial for headers, and size 18 font, Arial for the main text.

We also suggest using icons, many of which are freely available for noncommercial purposes, to emphasize certain points within the abstract. There are many icons which describe clinical elements, and are available from icon repositories such as Noun Project  and Flaticon . However, we caution authors that certain icons are licensed as “Creative Commons” and require attribution to the icon creator. However, many icons are licensed as either public domain (which do not require any attribution at all) or rights to use icons can be purchased for a nominal fee (generally <$10.00).

Lastly, we recommend using complementary color palettes when designing your abstract. Complementary color palettes improve readability and the visual appeal of your abstract. Many are freely available at Coolors.co . These color palettes can then be used to create backgrounds for specific information you wish to highlight in your abstract.

Once we incorporated all design elements, we developed the abstract below (Figure 2). However, we emphasize the need for multiple individuals to review and provide feedback on the abstract. Additionally, before posting the abstract to social media websites, we recommend exporting the abstract from Microsoft PowerPoint as a PNG file and reviewing it, to ensure that design elements remain consistent in the export process.

Figure 2. Finalized abstract.

Conclusions

Visual abstracts are an effective way of promoting research findings on social media, and for concisely communicating key findings to other researchers and the lay public. Within this blog post, we have provided some guidance on the creation of visual abstracts. We hope other members of the urologic community utilize this novel method of research outreach and communication.

Further Reading

NephJC Website —NephJC is a nephrology Twitter journal club that creates many high quality visual abstracts based on the nephrology literature. Their website contains many great examples of visual abstracts, and can be viewed for inspiration.

Andrew M. Ibrahim, MD , Creative Director at Annals of Surgery , provides an excellent guide on creating a visual abstract.

• Lindquist LA, Ramirez-Zohfeld V. Visual Abstracts to Disseminate Geriatrics Research Through Social Media. J Am Geriatric Soc. 2019;67(6):1128-1131.
• Ibrahim AM, Lillemoe KD, Klingensmith ME, Dimick JB. Visual Abstracts to Disseminate Research on Social Media: A Prospective, Case-control Crossover Study. Ann Surg. 2017;266(6):e46-e48.
• Zhu D, Srivastava A, Agalliu I, et al. Finasteride Use and Risk of Bladder Cancer in a Multiethnic Population. J Urol . in-press, doi:10.1097/JU.0000000000001694.

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

The PMC website is updating on October 15, 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• v.15(9); 2023 Sep
• PMC10590498

Graphical Abstract in Scientific Research

Madhan jeyaraman.

1 Orthopedics, ACS Medical College and Hospital, Dr MGR Educational and Research Institute, Chennai, IND

Harish V K Ratna

2 Orthopedics, Rathimed Speciality Hospital, Chennai, IND

Naveen Jeyaraman

Nicola maffulli.

3 Medicine, Surgery and Dentistry, University of Salerno, Baronissi, ITA

4 Orthopadic, Trauma, and Reconstructive Surgery, Rheinisch-Westfälische Technische Hochschule (RWTH) University Medical Centre, Aachen, DEU

5 School of Pharmacy and Bioengineering, Keele University Faculty of Medicine, Stoke-on-Trent, GBR

Filippo Migliorini

6 Centre for Sports and Exercise Medicine, Barts and the London School of Medicine and Dentistry, Mile End Hospital, Queen Mary University of London, London, GBR

Arulkumar Nallakumarasamy

Sankalp yadav.

7 Medicine, Shri Madan Lal Khurana Chest Clinic, New Delhi, IND

A graphical abstract (GA) summarizes the key and important findings of an article graphically, potentially stimulating researchers to view the published manuscript. A GA should enhance dissemination, augment engagement, and impact clinical practice. Infographics play a key role in a quicker understanding of the significant findings of a manuscript. Few level 1 studies reported that GAs enhanced the engagement of readers on social media when compared to plain text abstracts. With the evolution of Industry 4.0, 5.0, and 6.0, GA plays a major role in understanding the technical aspects of various technologies. This article outlines tips to prepare an effective GA and reports the impact of GAs on research and clinical translation.

Introduction and background

A graphical abstract (GA) is a graphic and visual summary of the important findings of a scientific article [ 1 ]. A GA should enhance a reader's ability to remember and recall information [ 2 ]. Recent research shows an eight-fold increase in the sharing of a visual abstract on social media in comparison to text-only abstracts; this has led to three times more visits by authors to the same article on the websites of journals [ 3 ]. In comparison to text abstracts, GAs have the upper hand: it takes about 6 seconds for an average reader to read about 20 words, whereas the same meaning could be understood within 1/4th of a second from a visual symbol [ 4 ].

Various social platforms, such as Facebook, Twitter, and Instagram, influence the lives of the general public and have become essential channels for communication among laypeople [ 5 ]. Among the total number of people using social media platforms, 1% are creators of content, 9% are editors, and the remaining 90% act as consumers of the content or seekers of the information provided to them [ 6 , 7 ]. In the present digital era, where there is exponential growth in the field of orthopedic research, GA plays a major role in helping readers preview the study content and decide whether to read the entire article [ 1 , 8 , 9 ].

GA enhances the dissemination of research among the readers, augments the engagement of research scholars, and creates a profound impact on clinical practice, which helps in clinical translation [ 1 ]. GA provides a bird's-eye view of the article, making it more attractive and appealing to readers, entices them to have an in-depth look, and makes it easier for the researcher to share or re-post on social media to engage like-minded individuals [ 10 ].

Triad of GA

The triad of GA is depicted in Figure ​ Figure1 1 .

Picture courtesy of Dr. Madhan Jeyaraman

The triad of GA is a) enhanced dissemination [GAs enhance the dissemination of research-related articles]. In the natural evolution of scientific information communication, videos and graphics will generate higher "hits" in comparison to text-only abstracts. Visual abstracts (VAs) increase the dissemination of research on social media by eight-fold compared to text-only abstracts [ 1 ]], b) augmented engagement [GAs provide a framework for increased engagement of like-minded researchers on social platforms. GAs act as a reminder to the reader and help in discussing the contents of the article in a forum], and c) impact on clinical practice [This may be difficult to evaluate, but, for example, the article published in "Annals of Surgery" had a great impact on disseminating information about antibiotic stewardship]. Producing an effective GA necessitated distilling the core research points into 2-3 eye-catching sentences to pique readers' interest . The prototype structure of GA is depicted in the graphical abstract of the manuscript. A GA can be divided into four areas, namely: a) the title area, b) the methods and cohort area, c) the findings area, and d) the conclusion area (Table ​ (Table1 1 ).

The tips and tricks for creating an effective GA are reported in Figure ​ Figure2 2 .

The design principles for a good GA are shown in Figure ​ Figure3 3 .

The elements of GA are: a) display of the title, name of the author, and name of the journal at the top of the GA; b) display of the journal’s logo at the bottom of the GA; c) display of the graphical presentation of research results; and d) display of the take-home message.

In general, GAs represent the face value of the underlying research article and are presented at the beginning of the article, after the text-only abstract. GAs provide a summary of the research article in the form of visual icons in single or dual-colored tones.

Resources to prepare GA

A social media platform enables the sharing of different GAs with various styles and structures. Peer-reviewed journals have started to provide guidelines for GAs based on recently published evidence, which would help in standardizing the outputs and would ensure the consistency and validity of the GAs published in a given journal. Many sources are available to provide guidelines for preparing a GA, which are as follows: a) Visual Abstract Primer (edited by Andrew Ibrahim), which covers topics such as creating a visual abstract and leveraging a visual abstract for dissemination [ 11 ], b) Andrew Ibrahim’s Guidelines to Standardize GAs for Scientific Research [ 1 ], and c) Michelle Lim’s short course on designing and the design process of GAs [ 12 ].

To prepare an informative GA effectively, one should be able to compile the research-related information and be able to reproduce it understandably under three main sections, namely a) the methods of the study, b) the important findings of the study, and c) the conclusions of the research study. There is no need for costly and complex illustrative software or additional graphical or artistic skills. The only additional need is good creativity to figure out the way to represent the findings in the form of visual icons, as well as the capacity to compile the information into small sections [ 12 ].

Process of creating and disseminating GA

The process of creating an effective GA is depicted in Figure ​ Figure4 4 .

Recently, more than 50 top journals, including "The New England Journal of Medicine", "Clinical Spine Surgery", and "JAMA", have adopted GAs to disseminate research through various social media platforms, particularly Twitter. By disseminating GAs through social media, over some time, the number of citations for the manuscript and the journal’s impact factor also steadily increased.

A few questions about GAs are being debated, namely: a) who can create GA, the author or the journal’s editorial team?; b) what are the specifications for creating GA to be included in the instruction for the author's section; c) who will review the GA (the author, the reviewer, the section editor, the editor-in-chief); d) who, how, and where will GAs be disseminated?; and e) if a journal has the answer to these questions, the dissemination of GAs among research scholars and clinicians will improve the quality of the GAs and maximize the impact of the research.

Cross-talk between GA and social media

In the recent past, there has been active participation by patients in surgical research, but their access to the results of the study in which they took part is limited to the commentary of non-expert individuals [ 13 , 14 ]. Research scientists must deliver the results of the research in a format easily accessible to readers. Various initiatives have been developed to enhance this communication and eradicate the communication gap, including the National Institute of Health Research's (NIHR) 'Make It Clear" campaign [ 15 ] and the British Medical Journal's ‘Patient and Public Partnership’ initiative [ 16 ]. Various social media platforms have become popular over the last 3-5 years to disseminate surgical research-related work [ 17 ]. This has been witnessed in growing instances, such as conference-related and specialty-related hashtags, facilitating discussion in international forums and journal-specific journal clubs, facilitating post-publication discussion among peers, and discussing new techniques of surgery through live surgical videos [ 17 , 18 ]. Another new initiative is GAs, which have helped to disseminate the results of surgical research to an academic audience [ 3 , 19 ].

GA can make an article more pleasant and understandable when compared to text-only tweets, allowing easy and fast recall of the important points of the article relevant to the research scholar [ 20 ]. The impact of GAs may rise or fall depending on the number of followers of the individual’s social platform account or may differ based on the particular specialty, subsequent sub-specialties, and social media platforms. The mean number of engagements on viewing visual abstracts (VAs) was significantly higher at 7 and 30 days than plain English texts, whereas the crossover results were similar for orthopedic research on social platforms, but greater overall public engagement was observed with visual abstracts than plain text tweets [ 20 ]. Some level 1 studies depict the usage of VAs through social media [ 3 , 21 , 22 ], with the strongest correlation between VAs and the dissemination of research on social media [ 3 ]. VAs may not increase the number of reads and downloads of the fully published manuscript [ 23 ]. These GAs make the article more appealing for the readers to grasp the main crux of that particular manuscript.

Challenges and pitfalls of GA

Any form of medium to exchange scientific research among readers has various challenges and pitfalls. Given the emergence of GA in various journals, the results of the underlying scientific research will be oversimplified, which may result in a misinterpretation of the results. When the readers systematically use GAs as a substitute for the whole manuscript, and this is not read, they may eventually lose the capability to evaluate and appraise manuscripts. When a GA is produced by a third party, the output of the GA may not be reviewed by the author and their team, which may theoretically introduce fallacies, misinterpretations, biases, or inaccuracies. The possibility of salami slicing may happen when the creation of GA is outsourced. The quality control of GAs remains a challenge when their production is outsourced. Though GA appears simple and lucid in understanding by the readers, the third party does not necessarily understand the subject to be able to convey the right message. The major setback for GA is the availability of space in the given area. So, authors tend to report only the positive findings of the study [ 12 ], inevitably introducing biases. GAs must be submitted with the main manuscript and undergo peer review.

Future directives

Future studies should focus on the impact of GAs on the citation of a scientific article. Among the available social media platforms, the best and ideal platform must be identified for promoting and propagating research among various orthopedic researchers. The journal may want to establish GA creation teams that have to undergo training in infographics, and proper feedback has to be obtained for sufficient training to produce GAs. The Nephrology Social Media Collective is a successful example of how to promote and propagate research through the creation of GA, podcasts, blogs, newsletters, and research games [ 24 , 25 ]. When the journal’s editorial team produces GAs, misinterpretation and bias, have to be minimized. However, the journal has to issue the guidelines and specifications for preparing GA to avoid the hassle in the editorial workflow. The creation of a GA repository within the journal will encourage readers to refer to the research results when needed [ 26 ]. The available way to search GA is through the hashtags #visualabstract or #graphicalabstract. GAs must have a link in the manuscript when searched via standard indexing and abstracting databases. An alternate form for searching GA is the creation of a GA repository after obtaining permission from the native journal for sharing and disseminating the GA.

GA can be presented at scientific conferences as a poster by using infographics and icons. This may attract more attention among conference attendees for the key findings of the research. GA can be presented as oral presentations in various forums, including clinical lectures, departmental and academic meetings, or case discussions, to project the research data more clearly. Finally, GA increases the visibility of the speaker and the research among like-minded research scholars. The GRAPHIC model must be adhered to to form a sound and effective GA, as depicted in Figure ​ Figure5 5 .

A schematic presentation of GA is depicted in Figure ​ Figure6 6 .

Conclusions

The emergence and integration of GAs into the landscape of scientific communication signifies an evolution in how research is presented and understood in our digital age. GAs, with their visual appeal and concise representation, are bridging the gap between dense scientific content and its audience, both experts and laypeople. Their popularity on social media platforms underscores the shift towards a more visual and immediate form of information consumption. However, like all innovations, GAs come with challenges, including potential misinterpretation and concerns about oversimplification. As we look forward, it's imperative for the scientific community to refine guidelines for creating and reviewing GAs, ensuring they both accurately represent the research and remain engaging to the audience. By doing so, we can harness the potential of GAs to improve the dissemination, engagement, and impact of scientific research without compromising the integrity of the content. This era of visual communication in science, championed by GAs, underscores the adage: a picture, indeed, might be worth a thousand words.

The authors have declared that no competing interests exist.

Author Contributions

Concept and design:   Sankalp Yadav, Madhan Jeyaraman, Harish V K. Ratna, Naveen Jeyaraman, Nicola Maffulli, Filippo Migliorini , Arulkumar Nallakumarasamy

Acquisition, analysis, or interpretation of data:   Sankalp Yadav, Madhan Jeyaraman, Harish V K. Ratna, Naveen Jeyaraman, Nicola Maffulli, Filippo Migliorini , Arulkumar Nallakumarasamy

Drafting of the manuscript:   Sankalp Yadav, Madhan Jeyaraman, Harish V K. Ratna, Naveen Jeyaraman, Nicola Maffulli, Filippo Migliorini , Arulkumar Nallakumarasamy

Critical review of the manuscript for important intellectual content:   Sankalp Yadav, Madhan Jeyaraman, Harish V K. Ratna, Naveen Jeyaraman, Nicola Maffulli, Filippo Migliorini , Arulkumar Nallakumarasamy

Supervision:   Sankalp Yadav, Naveen Jeyaraman, Nicola Maffulli, Filippo Migliorini

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings
• My Bibliography
• Collections
• Citation manager

Save citation to file

Email citation, add to collections.

• Create a new collection
• Add to an existing collection

Add to My Bibliography

Your saved search, create a file for external citation management software, your rss feed.

• Search in PubMed
• Search in NLM Catalog
• Add to Search

Visual Abstracts: Redesigning the Landscape of Research Dissemination

Affiliations.

• 1 Division of Nephrology and Hypertension, Department of Medicine, Vanderbilt University Medical Center, Nashville, TN.
• 2 Division of Nephrology and Hypertension, Department of Medicine, Vanderbilt University Medical Center, Nashville, TN. Electronic address: [email protected].
• PMID: 32560778
• DOI: 10.1016/j.semnephrol.2020.04.008

A visual abstract is a graphic summary of a study designed to enable readers to process key methods, findings, and conclusions rapidly. This allows readers to preview the article and decide if it is worth pursuing further. Similar to the text abstract, it is not a substitute for reading the full article. Its succinct format and attractive design make a visual abstract ideal for sharing on social media, thereby allowing journals and authors to promote published articles, and to facilitate discussion through tweets, blog posts, journal clubs, and scientific meetings. Guidelines for creating a visual abstract are available, but maintaining an acceptable standard remains a challenge. Visual abstract editors may be helpful in ensuring quality.

Keywords: Twitter; Visual abstract; medical education; research dissemination; social media.

Copyright © 2020 Elsevier Inc. All rights reserved.

PubMed Disclaimer

Similar articles

• A Picture Is Worth a Thousand Views: A Triple Crossover Trial of Visual Abstracts to Examine Their Impact on Research Dissemination. Oska S, Lerma E, Topf J. Oska S, et al. J Med Internet Res. 2020 Dec 4;22(12):e22327. doi: 10.2196/22327. J Med Internet Res. 2020. PMID: 33275112 Free PMC article.
• In a blink of an eye: Graphical abstracts in Advances in Clinical and Experimental Medicine. Misiak M, Kurpas D. Misiak M, et al. Adv Clin Exp Med. 2023 Sep;32(9):949-968. doi: 10.17219/acem/171615. Adv Clin Exp Med. 2023. PMID: 37676097
• Beyond Journals-Visual Abstracts Promote Wider Suicide Prevention Research Dissemination and Engagement: A Randomized Crossover Trial. Hoffberg AS, Huggins J, Cobb A, Forster JE, Bahraini N. Hoffberg AS, et al. Front Res Metr Anal. 2020 Oct 14;5:564193. doi: 10.3389/frma.2020.564193. eCollection 2020. Front Res Metr Anal. 2020. PMID: 33870046 Free PMC article.
• Globalization of continuing professional development by journal clubs via microblogging: a systematic review. Roberts MJ, Perera M, Lawrentschuk N, Romanic D, Papa N, Bolton D. Roberts MJ, et al. J Med Internet Res. 2015 Apr 23;17(4):e103. doi: 10.2196/jmir.4194. J Med Internet Res. 2015. PMID: 25908092 Free PMC article. Review.
• Creating Effective Infographics and Visual Abstracts to Disseminate Research and Facilitate Medical Education on Social Media. Spicer JO, Coleman CG. Spicer JO, et al. Clin Infect Dis. 2022 May 15;74(Suppl_3):e14-e22. doi: 10.1093/cid/ciac058. Clin Infect Dis. 2022. PMID: 35568482 Review.
• Harnessing Social Media to Enhance Nephrology Academia. Meena P, Mohanasundaram S, Kurian J, Prasad GS, Bhargava V, Panda S, Agrawaal KK. Meena P, et al. JNMA J Nepal Med Assoc. 2023 Sep 1;61(265):741-747. doi: 10.31729/jnma.8268. JNMA J Nepal Med Assoc. 2023. PMID: 38289794 Free PMC article. Review.
• Embracing the (r)evolution of social media and digital scholarship in pediatric nephrology education. Shah SS, Zangla E, Qader MA, Chaturvedi S, Mannemuddhu SS. Shah SS, et al. Pediatr Nephrol. 2024 Jul;39(7):2061-2077. doi: 10.1007/s00467-023-06251-y. Epub 2023 Dec 27. Pediatr Nephrol. 2024. PMID: 38150027 Review.
• Seeing the unseen: Comparison study of representation approaches for biochemical processes in education. Pokojná H, Kozlíková B, Berry D, Kriglstein S, Furmanová K. Pokojná H, et al. PLoS One. 2023 Nov 6;18(11):e0293592. doi: 10.1371/journal.pone.0293592. eCollection 2023. PLoS One. 2023. PMID: 37930950 Free PMC article. Review.
• Racial, Gender, and Size Bias in a Medical Graphical Abstract Gallery: A Content Analysis. Cerdeña JP, Tsai JW, Warpinski C, Rosencrans RF, Gravlee CC. Cerdeña JP, et al. Health Equity. 2023 Sep 27;7(1):631-643. doi: 10.1089/heq.2023.0026. eCollection 2023. Health Equity. 2023. PMID: 37786527 Free PMC article.
• Guidelines: The Do's, Don'ts and Don't Knows of Creating Open Educational Resources. Khalid F, Wu M, Ting DK, Thoma B, Haas MRC, Brenner MJ, Yilmaz Y, Kim YM, Chan TM. Khalid F, et al. Perspect Med Educ. 2023 Jan 9;12(1):25-40. doi: 10.5334/pme.817. eCollection 2023. Perspect Med Educ. 2023. PMID: 36908747 Free PMC article.

Publication types

• Search in MeSH

LinkOut - more resources

Full text sources.

• ClinicalKey
• Elsevier Science
• Citation Manager

NCBI Literature Resources

MeSH PMC Bookshelf Disclaimer

The PubMed wordmark and PubMed logo are registered trademarks of the U.S. Department of Health and Human Services (HHS). Unauthorized use of these marks is strictly prohibited.

Unfortunately we don't fully support your browser. If you have the option to, please upgrade to a newer version or use Mozilla Firefox , Microsoft Edge , Google Chrome , or Safari 14 or newer. If you are unable to, and need support, please send us your feedback .

We'd appreciate your feedback. Tell us what you think!   opens in new tab/window

Picture this: a stimulating way of opening up your research to new audiences

April 19, 2021 | 5 min read

By Christopher Tancock

How visual abstracts can help you get noticed

As with many things in life, the process of submitting and publishing an article can seem like climbing a multi-levelled hill (or mountain in some cases!). Producing a draft gets you to one level; navigating the submission process is another uphill slog. Climbing and conquering the heights of peer review, revision and finally, with luck, acceptance requires a host of patience and a mastery of different techniques. You might think this would be the end of the journey, but having your article accepted is in many ways just another beginning. Look around you – there are quite a few other authors standing around at this level – all hoping their article gets noticed by a broad audience. So to reach that final level and truly bask in the warm sunlight of success, there’s one more step to take: we need to paint a picture.

A great way to stand out and ensure your work reaches the broadest audience is to make your work as  accessible  as possible and in an age where everyone’s time is precious (and scarce), it’s useful to remember that old adage: “a picture is worth a thousand words”. Enter the visual/graphical abstract (VGA) * .

The visual abstract was originally the brainchild of  Professor Andrew Ibrahim   opens in new tab/window , the Creative Director of  Annals of Surgery . From occasional appearances in a single journal, VGAs have now become much more common and you’ll find many journals now mandating their submission alongside the main body of the article. At Elsevier alone, we saw over 130,000 graphical abstracts published in 2020, a 33% increase on the number in 2019. Around 1,400 of our journals have published at least one VGA in the last year. But what  is  a visual abstract?

A visual abstract is a single, concise, pictorial and visual summary of the main findings of the article. A VGA allows readers to quickly gain an understanding of the take-home message of the paper and is intended to encourage browsing, promote interdisciplinary scholarship, and help readers identify more quickly which papers are most relevant to their research interests.

There are various ways to approach constructing a graphical abstract (and note that many journals have their own guidelines and even templates for doing so). Elsevier’s Author Hub has  a page dedicated to VGAs  (along with instructions, best practices and examples of good VGAs). There are also some excellent resources elsewhere including  Professor Ibrahim’s fantastic Primer   opens in new tab/window , now in its fourth edition.

Regardless of the specifics, a VGA is usually divided into sections and the whole piece should tell a story, encouraging the viewer to want to read the full article. Here are our top tips for drawing up your abstract:

Think about the purpose/audience for your abstract.

Use detail sparingly without oversimplifying or misrepresenting.

Remember to include a citation back to the article – perhaps as a QR code in addition to the DOI/URL.

Don’t clutter the abstract: white space is your friend here!

Make good use of design and infographical elements e.g. icons and pictograms.

Be creative when designing your VGA and bear these key descriptors in mind: “simplify”, “illustrate”, “explain”, “entice”.

Examples of visual/graphical abstracts

Sleep Apnea in Maintenance Hemodialysis: A Mixed-Methods Study   opens in new tab/window

A meta-research study revealed several challenges in obtaining placebos for investigator-initiated drug trials   opens in new tab/window

Naturally such a construct lends itself superbly to dissemination on social media as well as use in online “journal club” discussions, in blog posts, at conferences and in presentations. You could even get a smaller version printed out to disseminate at events and meetings. If you invest some time, you can use VGAs to help make a splash with your article, often with impressive results. Research has shown that graphical abstracts impact positively on their articles both in terms of views of the article as well as increased activity on social media. In particular,  the average annual use of an article is doubled when compared with those without a visual abstract   opens in new tab/window .

If you know that your research might be of interest to different audiences (say, fellow researchers and the general public), it could be a wise idea to construct a different VGA for each audience. You can further increase the accessibility of your work by combining the graphical abstract with a  lay summary . (This wouldn’t be necessary for every article but key research you want to open up should have its own “campaign”, designed to encourage your peers to read the full article whilst also educating the general public about what you’ve done.)

In summary, visual/graphical abstracts are a fantastic device to showcase the essential thrust of your research and, used properly, can help break down barriers and open your work up to new audiences as well as ensuring more usage. It’s important to employ a visual abstract as part of a wider strategy to promote your work including use of good keywords, social media promotion, use of  Share Links  and similar tools, and the preparation of a lay summary. What VGAs are  not  is a substitute for reading the (whole) article – view them more as a  gateway  to said work.

Given the rapid expansion of submitted and published graphical abstracts at Elsevier alone, it’s reasonable to conclude that they will play a significant and growing part in the academic publication process. Questions have been raised about whether VGAs should be formalized and integrated into the peer review process to ensure quality. We’ll continue to monitor trends, offer advice and help you to get the most out of your work. In the meantime we’ll leave you with this offering: “a picture is worth 1,000 words; a  good  picture 10,000!”.

* The terms  visual- and graphical abstract are used interchangeably in this piece reflecting a widespread use of both alternatives in the community.  We are currently considering which terminology to employ going forward. Have an opinion on this?  Let us know via the short poll below!

Contributor

Christopher Tancock

Authors' update - keeping journal authors in touch with industry developments, support and training.

• - Google Chrome

Intended for healthcare professionals

• My email alerts
• BMA member login
• Username * Password * Forgot your log in details? Need to activate BMA Member Log In Log in via OpenAthens Log in via your institution

Search form

• Advanced search
• Search responses
• Search blogs

BMJ visual abstracts

To help our readers to get a quick overview of research we publish, we started making visual abstracts (also known as "graphical abstracts") in March 2018. These small images give a summary of selected papers. Initially, we are focussing on reports of trials and systematic reviews, but more formats may be introduced in future.

7 August 2024

24 July 2024

19 June 2024

10 April 2024

28 February 2024

14 February 2024

7 February 2024

24 January 2024

15 December 2023

31 October 2023

27 September 2023

2 August 2023

11 July 2023

13 June 2023

16 May 2023

19 April 2023

29 March 2023

22 February 2023

25 January 2023

21 December 2022

30 November 2022

23 November 2022

1 November 2022

28 September 2022

21 September 2022

17 August 2022

30 June 2022

13 June 2022

25 May 2022

6 April 2022

9 March 2022

2 February 2022

12 January 2022

16 December 2021

14 December 2021

18 November 2021

14 October 2021

27 August 2021

07 July 2021

24 June 2021

13 May 2021

14 April 2021

17 February 2021

13 January 2021

16 December 2020

4 November 2020

7 October 2020

30 September 2020

18 August 2020

01 July 2020

14 May 2020

13 May 2020

1 April 2020

11 March 2020

25 February 2020

22 January 2020

18 December 2019

5 November 2019

16 October 2019

18 September 2019

24 July 2019

05 June 2019

17 April 2019

09 April 2019

27 March 2019

13 March 2019

20 February 2019

13 February 2019

30 January 2019

23 January 2019

10 December 2018

31 October 2018

10 October 2018

26 September 2018

12 September 2018

04 September 2018

29 August 2018

19 July 2018

19 June 2018

13 June 2018

17 May 2018

16 May 2018

09 May 2018

03 May 2018

02 May 2018

21 March 2018

• International jobs

This week's poll

Read related article

See previous polls

Visual Abstracts

What is a visual abstract.

A visual abstract is a visual summary of the key findings of an article. Like the abstract section of an article; it conveys the most essential points in a shorter format, but it does not replace reading the full article. Instead, it serves to generate reader interest.

How does a visual abstract help my article?

In today’s digital environment, you have just a few seconds to capture the audience’s attention. High-impact visuals are just one of the tools that help to achieve this. A 2017 study by Ibrahim et al showed that, compared to text-only tweets promoting a published article, tweets with a visual abstract had 7-fold higher impressions, 8-fold higher retweets, and nearly 3-fold higher article visits on the publisher website (1). Other trials have shown visual abstracts to be similarly effective across social media (2,3).

Creating a visual abstract for your PCD manuscript

During the peer review process, authors may be invited by PCD’s Editor in Chief to create and submit a visual abstract for consideration along with their manuscript. If a visual abstract is requested, the corresponding author will receive an email invitation to work with the PCD team to make stylistic edits and approve a final version for release. The manuscript and visual abstract are reviewed independently, so a visual abstract might be rejected even though the accompanying manuscript is accepted for publication. The corresponding author will serve as the single point of contact and will be responsible for collecting and combining comments from coauthors.

PCD’s best practices for visual abstracts:

• Identify 1-3 key points or outcomes in the manuscript.
• Build a slide with one panel for each key point using PCD’s template for visual abstracts below. You may have 1-3 panels in a slide.
• Enter the title, first author name, and key points into the template.
• Add visuals to convey each point. Be sure to use ONLY images and graphics that are original or are within the public domain. Copyrighted images will not be accepted.
• Save your slide in Microsoft Word or PowerPoint and submit it as a supplementary file along with all other manuscript documents at https://mc.manuscriptcentral.com/pcd external icon .

PCD Template

For examples please see PCD’s Visual Abstracts gallery.

• Ibrahim AM, Lillemoe KD, Klingensmith ME, Dimick JB. Visual Abstracts to Disseminate Research on Social Media: A Prospective, Case-control Crossover Study. Ann Surg 2017;266(6):e46-e48.
• Lindquist LA, Ramirez-Zohfeld V. Visual Abstracts to disseminate geriatrics research through social media. J Am Geriatr Soc 2019;67(6):1128-1131.
• Koo K, Aro T, Pierorazio PM. Impact of Social Media Visual Abstracts on Research Engagement and Dissemination in Urology. J Urol 2019;202(5):875-877.

Additional reading

• https://www.surgeryredesign.com/resources external icon
• Search #VisualAbstract on Twitter for more examples

The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the U.S. Department of Health and Human Services, the Public Health Service, the Centers for Disease Control and Prevention, or the authors’ affiliated institutions.

To receive email updates about this page, enter your email address:

• PCD on Instagram
• Get LinkedIn
• PCD on Twitter
• PCD on Facebook
• PCD Sound Bites
• Subscribe to RSS
• PubMed Central

• Emerging Infectious Diseases Journal
• Morbidity and Mortality Weekly Report (MMWR)
• Vital Signs

212-460-1500

[email protected]

Visual Abstracts are eye-catching, graphical summaries of your research. They are very effective at engaging your audience on social media channels.

Visual Abstract

Regular price $190.00.

Unit price / per 

Name of publication's first author*

Abstract text* You can add text upto 600 characters. 600 character(s) remaining. -->

Abstract/Publication title*

DOI number (if not available please type 'NA')*

Journal citation (if not available please type 'NA')*

Notes to production team

I have read and accepted the Terms of Use and Privacy Statement.

What are Visual Abstracts?

Scope We cover all major scientific disciplines*. 

How can Visual Abstracts help you?

• Make an impact

Make a strong impact and increase your article views with Visual Abstracts optimized for social media channels.

• Draw attention

Highlight your work with illustrations designed by a team with decades of experience in science illustration.

• Communicate well

Communicate your findings to a broad audience and in a fraction of time. Our Visual Abstracts are crafted to increase readability.

What makes our product best for you ?

Researchers from all over the world helped us to develop and to test this unique product: Our Visual Abstracts were engineered to guide the reader's eye and to increase reading speed without reducing content memorization.

How to use your Visual Abstract on Social Media?

What is a Graphical Abstract and Why Do I Need One for My Paper?

Frequently Asked Questions

*Some areas like theoretical physics and mathematics are not yet amenable to our visualization service. Only abstracts written in English language can be processed at this stage. We are working hard to expand our service to other scientific disciplines and other languages.

Use this popup to embed a mailing list sign up form. Alternatively use it as a simple call to action with a link to a product or a page.

Age verification

By clicking enter you are verifying that you are old enough to consume alcohol.

Search solutions

Your cart is currently empty..

Visual Abstracts

Visual abstracts are a visual way, using illustrations, icons and flow-chart graphics to present the abstract of your research paper. They tend to provide this information in a single “slide”, relying more on imagery than on text to communicate the overall goal and outcomes of your research.

UF Health Communications has prepared a series of templates that will provide a starting point from which you can customize your own visual abstracts for publisher submission, social media posts and more. Note: if using these on social media, please ensure you provide a detailed alternative text (a written summary of the visual and text components), so those utilizing assistive devices are able to understand what content is being shared.

Each template contains an example along with a generic version you can use to integrate your own data.

Visual Abstract

Template #01

Consider this 3-panel format to present three salient points.

VISUAL ABSTRACT

Template #02

Consider this 6-panel format to present context and/or intervention design with additional data.

Template #03

Consider this format to present studies that show an “A vs. B” scenario. For instance, this template could be used to present “results vs. control” findings or “Drug A vs. Drug B” research.

Template #04

Consider this three-part narrative format to present an overview of hypothesis, methodology and outcome, with summarization/key findings highlighted.

Template #05

Consider this format to visually depict “A vs. B” treatments, while highlighting the research goal(s).

Graphic Resources

Libraries of royalty-free and Creative Commons licensed icons, illustrations, and other graphics.

• BioRender – create science figures
• Reactome – Library of icons for Enhanced High Level Diagrams (EHLD)
• The Noun Project – a large collection of general icons
• Bioicons – biology-based icons and illustrations
• SciDraw – free repository of high quality drawings useful for scientific presentations
• Servier Medical Art – 3,000 free medical images
• CDC – Public Health Image Library (PHIL)
• National Cancer Institute – Visuals Online

Visual Abstract Resources

Additional examples and guidance on how to create effective visual abstracts

• Visual Abstract Primer
• Elsevier – Graphical Abstracts
• The Annals of Thoracic Surgery – Visual Abstracts
• Arcia A, Suero-Tejeda N, Bales ME, et al. Sometimes more is more: iterative participatory design of infographics for engagement of community members with varying levels of health literacy. J Am Med Inform Assoc. 2016;23:174-183.
• Atkinson C. Beyond Bullet Points, 4th ed. London: Pearson Education, 2018.
• Duarte N. Slide:ology. The Art and Science of Creating Great Presentations. Sebastopol, CA: O’Reilly Media, Inc, 2008.
• Duarte N. Resonate. Present Visual Stories that Transform Audiences. Hoboken, NJ: John Wiley & Sons, Inc, 2010.
• Ibrahim AM (ed). Use of a Visual Abstract to Disseminate Scientific Research , Version 4, January 2018.
• Ibrahim A, Lillemoe KD, Klingensmith ME, Dimick JB. Visual abstracts to disseminate research on social media: a prospective, case-control crossover study. Ann Surg. 2017;266(6):e46-e48.
• Lam J. When less is more: writing great copy for visual content .
• Lang T. Up and down or side by side: structuring comparisons in data tables. AMWA J. 2018;33(3):104-110.
• Levie WH, Lentz R. Effects of text illustrations: a review of research. Ed Comm Tech J. 1982;30(4):195-232.
• Martin LJ, Turnquist A, Groot B, et al. Exploring the role of infographics for summarizing medical literature. Health Prof Educ. 2018 March [in press].
• National Cancer Institute. Making Data Talk: A Workbook .
• Pernice K. F-shaped pattern of reading on the web: misunderstood but still relevant (even on mobile) .
• University of Michigan. Visualizing health. A scientifically vetted style guide for communicating health data .
• Vogel DR, Dickson GW, Lehman JA. Persuasion and the role of visual presentation support: the UM/3M study . 1986.
• Weinreich H, Obendorf H, Herder E, Mayer M. Not quite the average: an empirical study of web use. ACM Transactions on the Web.2008;2(1):article 5.

• Introduction
• Conclusions
• Article Information

The frequency that each abstract version was ranked (first, second, third, and fourth) by surgeon-reviewers. Chatbot 1 refers to Chat Generative Pretrained Transformer (GPT) version 3.5; chatbot 2, Chat-GPT version 4.0.

eAppendix 1. Chat GPT Training and Writing Prompts

eAppendix 2. 10-Point and 20-Point Scale Rubrics

eAppendix 3. Chat GPT Grading Prompts

Data Sharing Statement

See More About

Sign up for emails based on your interests, select your interests.

Customize your JAMA Network experience by selecting one or more topics from the list below.

• Academic Medicine
• Acid Base, Electrolytes, Fluids
• Allergy and Clinical Immunology
• American Indian or Alaska Natives
• Anesthesiology
• Anticoagulation
• Art and Images in Psychiatry
• Artificial Intelligence
• Assisted Reproduction
• Bleeding and Transfusion
• Caring for the Critically Ill Patient
• Challenges in Clinical Electrocardiography
• Climate and Health
• Climate Change
• Clinical Challenge
• Clinical Decision Support
• Clinical Implications of Basic Neuroscience
• Clinical Pharmacy and Pharmacology
• Complementary and Alternative Medicine
• Consensus Statements
• Coronavirus (COVID-19)
• Critical Care Medicine
• Cultural Competency
• Dental Medicine
• Dermatology
• Diabetes and Endocrinology
• Diagnostic Test Interpretation
• Drug Development
• Electronic Health Records
• Emergency Medicine
• End of Life, Hospice, Palliative Care
• Environmental Health
• Equity, Diversity, and Inclusion
• Facial Plastic Surgery
• Gastroenterology and Hepatology
• Genetics and Genomics
• Genomics and Precision Health
• Global Health
• Guide to Statistics and Methods
• Hair Disorders
• Health Care Delivery Models
• Health Care Economics, Insurance, Payment
• Health Care Quality
• Health Care Reform
• Health Care Safety
• Health Care Workforce
• Health Disparities
• Health Inequities
• Health Policy
• Health Systems Science
• History of Medicine
• Hypertension
• Images in Neurology
• Implementation Science
• Infectious Diseases
• Innovations in Health Care Delivery
• JAMA Infographic
• Law and Medicine
• Leading Change
• Less is More
• LGBTQIA Medicine
• Lifestyle Behaviors
• Medical Coding
• Medical Devices and Equipment
• Medical Education
• Medical Education and Training
• Medical Journals and Publishing
• Mobile Health and Telemedicine
• Narrative Medicine
• Neuroscience and Psychiatry
• Notable Notes
• Nutrition, Obesity, Exercise
• Obstetrics and Gynecology
• Occupational Health
• Ophthalmology
• Orthopedics
• Otolaryngology
• Pain Medicine
• Palliative Care
• Pathology and Laboratory Medicine
• Patient Care
• Patient Information
• Performance Improvement
• Performance Measures
• Perioperative Care and Consultation
• Pharmacoeconomics
• Pharmacoepidemiology
• Pharmacogenetics
• Pharmacy and Clinical Pharmacology
• Physical Medicine and Rehabilitation
• Physical Therapy
• Physician Leadership
• Population Health
• Primary Care
• Professional Well-being
• Professionalism
• Psychiatry and Behavioral Health
• Public Health
• Pulmonary Medicine
• Regulatory Agencies
• Reproductive Health
• Research, Methods, Statistics
• Resuscitation
• Rheumatology
• Risk Management
• Scientific Discovery and the Future of Medicine
• Shared Decision Making and Communication
• Sleep Medicine
• Sports Medicine
• Stem Cell Transplantation
• Substance Use and Addiction Medicine
• Surgical Innovation
• Surgical Pearls
• Teachable Moment
• Technology and Finance
• The Art of JAMA
• The Arts and Medicine
• The Rational Clinical Examination
• Tobacco and e-Cigarettes
• Translational Medicine
• Trauma and Injury
• Treatment Adherence
• Ultrasonography
• Users' Guide to the Medical Literature
• Vaccination
• Venous Thromboembolism
• Veterans Health
• Women's Health
• Workflow and Process
• Wound Care, Infection, Healing

Get the latest research based on your areas of interest.

Others also liked.

• Download PDF
• X Facebook More LinkedIn

Holland AM , Lorenz WR , Cavanagh JC, et al. Comparison of Medical Research Abstracts Written by Surgical Trainees and Senior Surgeons or Generated by Large Language Models. JAMA Netw Open. 2024;7(8):e2425373. doi:10.1001/jamanetworkopen.2024.25373

Manage citations:

© 2024

• Permissions

Comparison of Medical Research Abstracts Written by Surgical Trainees and Senior Surgeons or Generated by Large Language Models

• 1 Division of Gastrointestinal and Minimally Invasive Surgery, Department of Surgery, Atrium Health Carolinas Medical Center, Charlotte, North Carolina
• 2 Department of Economics, Massachusetts Institute of Technology, Cambridge
• 3 Division of Colorectal Surgery, Department of Surgery, Royal Devon & Exeter Hospital, Exeter, Devon, United Kingdom
• 4 Department of Clinical Medicine, University of Copenhagen, Bispedjerg & Frederiksberg Hospital, Copenhagen, Denmark
• 5 Division of Plastic and Reconstructive Surgery, The Ohio State University Wexner Medical Center, Columbus
• 6 Division of Plastic Surgery, University of Pennsylvania Health System, Philadelphia

Question   Can large language models generate convincing medical research abstracts?

Findings   In this cross-sectional study comparing 10 medical abstracts written by surgical trainees and senior surgeons or generated by large language models, blinded expert surgeon-reviewers were asked to grade and rank these abstracts. There was no statistical difference in the grades or ranks of abstracts generated by the language model when compared with abstracts written by surgical trainees or senior surgeons.

Meaning   These findings suggest that when appropriately trained with background literature, abstract formatting, primary research data, and a thorough prompt, chatbots can generate medical research abstracts that are difficult to distinguish from surgeon-scientist–written abstracts.

Importance   Artificial intelligence (AI) has permeated academia, especially OpenAI Chat Generative Pretrained Transformer (ChatGPT), a large language model. However, little has been reported on its use in medical research.

Objective   To assess a chatbot’s capability to generate and grade medical research abstracts.

Design, Setting, and Participants   In this cross-sectional study, ChatGPT versions 3.5 and 4.0 (referred to as chatbot 1 and chatbot 2) were coached to generate 10 abstracts by providing background literature, prompts, analyzed data for each topic, and 10 previously presented, unassociated abstracts to serve as models. The study was conducted between August 2023 and February 2024 (including data analysis).

Exposure   Abstract versions utilizing the same topic and data were written by a surgical trainee or a senior physician or generated by chatbot 1 and chatbot 2 for comparison. The 10 training abstracts were written by 8 surgical residents or fellows, edited by the same senior surgeon, at a high-volume hospital in the Southeastern US with an emphasis on outcomes-based research. Abstract comparison was then based on 10 abstracts written by 5 surgical trainees within the first 6 months of their research year, edited by the same senior author.

Main Outcomes and Measures   The primary outcome measurements were the abstract grades using 10- and 20-point scales and ranks (first to fourth). Abstract versions by chatbot 1, chatbot 2, junior residents, and the senior author were compared and judged by blinded surgeon-reviewers as well as both chatbot models. Five academic attending surgeons from Denmark, the UK, and the US, with extensive experience in surgical organizations, research, and abstract evaluation served as reviewers.

Results   Surgeon-reviewers were unable to differentiate between abstract versions. Each reviewer ranked an AI-generated version first at least once. Abstracts demonstrated no difference in their median (IQR) 10-point scores (resident, 7.0 [6.0-8.0]; senior author, 7.0 [6.0-8.0]; chatbot 1, 7.0 [6.0-8.0]; chatbot 2, 7.0 [6.0-8.0]; P  = .61), 20-point scores (resident, 14.0 [12.0-7.0]; senior author, 15.0 [13.0-17.0]; chatbot 1, 14.0 [12.0-16.0]; chatbot 2, 14.0 [13.0-16.0]; P  = .50), or rank (resident, 3.0 [1.0-4.0]; senior author, 2.0 [1.0-4.0]; chatbot 1, 3.0 [2.0-4.0]; chatbot 2, 2.0 [1.0-3.0]; P  = .14). The abstract grades given by chatbot 1 were comparable to the surgeon-reviewers’ grades. However, chatbot 2 graded more favorably than the surgeon-reviewers and chatbot 1. Median (IQR) chatbot 2-reviewer grades were higher than surgeon-reviewer grades of all 4 abstract versions (resident, 14.0 [12.0-17.0] vs 16.9 [16.0-17.5]; P  = .02; senior author, 15.0 [13.0-17.0] vs 17.0 [16.5-18.0]; P  = .03; chatbot 1, 14.0 [12.0-16.0] vs 17.8 [17.5-18.5]; P  = .002; chatbot 2, 14.0 [13.0-16.0] vs 16.8 [14.5-18.0]; P  = .04). When comparing the grades of the 2 chatbots, chatbot 2 gave higher median (IQR) grades for abstracts than chatbot 1 (resident, 14.0 [13.0-15.0] vs 16.9 [16.0-17.5]; P  = .003; senior author, 13.5 [13.0-15.5] vs 17.0 [16.5-18.0]; P  = .004; chatbot 1, 14.5 [13.0-15.0] vs 17.8 [17.5-18.5]; P  = .003; chatbot 2, 14.0 [13.0-15.0] vs 16.8 [14.5-18.0]; P  = .01).

Conclusions and Relevance   In this cross-sectional study, trained chatbots generated convincing medical abstracts, undifferentiable from resident or senior author drafts. Chatbot 1 graded abstracts similarly to surgeon-reviewers, while chatbot 2 was less stringent. These findings may assist surgeon-scientists in successfully implementing AI in medical research.

The introduction of artificial intelligence (AI) into the medical field has been both a promising and polarizing venture. Particularly, OpenAI Chat Generative Pretrained Transformer (ChatGPT; versions 3.5 and 4.0) is a new large language model, or chatbot, that has been trained from massive datasets to respond to prompts with sophisticated human-like answers. 1 , 2 Medical professionals agree that these large language models have opened the door for new possibilities in medicine but also Pandora’s box. Arguments can be made for the benefit of AI in scientific research as well as for conflicts associated with AI in medicine.

The most common controversies associated with chatbots are the encroachment of plagiarism, biased training data, lack of creativity, and the spread of misinformation. 3 Many surgeon-scientists worry that chatbots pull from sources that cannot be given proper credit, leading to plagiarism and copyright infringement. 4 , 5 Although chatbots are trained on a plethora of information, there is little transparency in the data’s origin. 1 , 6 , 7 As new reporting guidelines 7 - 9 recommend how to describe the role of AI in a project, publishers and editors grapple with the listing of chatbots as an author. Some argue that chatbots should not be listed as an author because they cannot take responsibility for what is written. 1 , 7 , 10 - 12 The ability of chatbots to generate novel ideas or think critically has also been questioned. 4 , 13 - 15 Of particular concern is the spread of misinformation. 4 , 10 Chatbots are not trained exclusively on medical texts, so there can be blatant inaccuracies (ie, hallucinations) in some of the AI responses. 2 , 16 - 18 Chatbots state this information with a false confidence that precludes inaccuracy unless scrutinized by a well-versed health care clinician. 18 Whether chatbots are endorsed by the scientific community or not, patients will inevitably use them to answer medical questions, so physicians should be invested in how to best validate the knowledge they emit. 4 , 15 , 18

As a counterargument to these concerns, AI has several beneficial applications to the field of health care. 1 , 4 , 7 , 18 - 21 Chatbots have demonstrated the ability to translate text 4 , 11 and be integrated into hospital electronic medical records. 21 They have even passed the US Medical Licensing Examination steps 1 and 2, which are required by medical students to earn their degree. 22 The role of chatbots in scientific writing is being explored 23 - 25 with the goal of improving efficiency and productivity of surgeon-scientists. 4 , 6 , 10 , 14 If chatbots can be trained to assist in generating text for publication, scientists can devote more time to the complex pursuits involved in research. 1 , 2 , 4 The goal of our study was to train 2 chatbots to generate medical research abstracts and assess how these abstracts compared with resident- and senior author–written abstracts as reviewed by blinded, well-published surgeons in the field. Furthermore, we evaluated the ability of chatbots to grade and rank medical abstracts when taught with a rubric.

This cross-sectional study was performed at a tertiary care center in the Southeastern US and was determined exempt from review and the requirement of informed consent by the Carolinas Medical Center institutional review board. All abstracts utilized were written about a study previously approved by the Carolinas Medical Center institutional review board. This report follows Strengthening the Reporting of Observational Studies in Epidemiology ( STROBE ) reporting guideline. The study was conducted between August 2023 and February 2024 (including data analysis).

OpenAI ChatGPT (versions 3.5 and 4.0; hereafter referred to as chatbot 1 and chatbot 2) was trained to generate medical abstracts based on provided abstracts as examples. The research residents and senior attending physician identified 10 abstracts 26 - 35 by our group from 2012 to 2022 that were presented at national meetings and published in surgical journals to serve as the training models. There was variation in the first author of each abstract, a junior trainee, but all studies had the same senior author (B.T.H.). These abstracts were inputted as examples of our group’s writing style to provide few-shot learning (training an existing model by providing it examples to work from) for chatbot 1 and chatbot 2. The chatbots were prompted to note the similarities between the abstracts and confirm that they had saved our writing style. See eAppendix 1 in Supplement 1 for exact prompts.

Ten additional abstracts 36 - 45 were used to investigate the chatbots’ ability to generate scientific abstracts. These abstracts were written by 5 different trainees within the first 6 months of their research year at the same medical center between 2018 to 2023 to account for the novice period. Abstracts from the current year’s research residents and fellows were excluded. All abstracts had the same senior author as the training abstracts (B.T.H.) and were submitted and presented at a variety of national and international conferences. Finally, these abstracts could only be included if we had access to the initial draft and final submitted version, the statistically analyzed research data, and a literature review of information concerning the topic of the abstract.

Once the chatbots were trained, we asked that it generate a scientific abstract based on the information provided. For each of the 10 abstracts, the chatbots were given the introduction and discussion of 3 relevant publications. 46 - 75 Text limitations prevented us from giving the chatbots the entire article. Next, we provided our prompt. 6 , 16 Specifically, we told the chatbots to generate text in the style of a senior surgeon-scientist with over 20 years of experience, like our senior author (B.T.H.). The analyzed real-world research data from each study was then pasted into the chat box. Finally, using the background literature, its knowledge as an experienced surgeon, and the data analysis, we asked both chatbot 1 and chatbot 2 to generate a version of each abstract in the trained writing style and in the specified format that was required by each national conference. An example prompt is available in eAppendix 1 in Supplement 1 .

Once chatbot 1 and chatbot 2 generated abstracts of each of the 10 studies, these were compared with the resident’s first unedited draft and the senior surgeon’s edited, submitted version of the same abstract. The 4 versions were deidentified and sent to 5 blinded surgeon-reviewers (J.E.J., L.N.J, J.P.F., N.J.S., and K.W.K.). The 5 surgeons come from academic practices in Denmark, the UK, and the US, and all have served as presidents or board members of international surgical organizations or editorial boards with extensive experience in research and abstract writing and grading. The reviewers were asked to independently score the 4 versions of the abstracts on a 10- and 20-point scale. The 10-point scale was based on a typical abstract rubric. The 20-point scale was based on the American Society of Plastic Surgeons, which entailed 4 categories: completeness, relevance, quality, and exposure (each worth 5 points). See eAppendix 2 in Supplement 1 for the rubrics. The reviewers were also asked to force rank the 4 abstract versions from first to fourth, with first being the best abstract and fourth being the worst, with no ties. They were asked to repeat these grading methods for all 10 abstracts for a total of 40 versions. Additionally, in a separate session, we tasked chatbot 1 and chatbot 2 with grading all 40 abstract versions. The chatbots were provided with the same instructions on a standard 10-point rubric with 10 being the best and a 20-point rubric broken into 4 categories: completeness, relevance, quality, and exposure. See eAppendix 3 in Supplement 1 for the prompt and rubric provided to the chatbots.

Standard descriptive and comparison statistics were performed on the abstract versions using SAS version 9.4 (SAS Institute). The Fisher exact test was applied to compare categorical variables, and Kruskal-Wallis was utilized to compare continuous variables. All P values were 2-sided, and statistical significance was set at P  < .05. We hypothesized that the chatbots would generate similarly graded and ranked abstracts as those written by surgical trainees and senior surgeons.

Each surgeon-reviewer ranked an AI-generated version of an abstract first at least once, and 1 reviewer ranked either the chatbot 1 or chatbot 2 version first every time. The surgeon-reviewers ranked the resident’s version first 14 of 50 times and last 14 of 50 times. They ranked the senior author’s version first 13 of 50 times and last 13 of 50 times. The chatbot 1 version was ranked first least often (7 of 50 times) and ranked last most often (16 of 50 times). The chatbot 2 version was ranked first most often (16 of 50 times) and was ranked last least often (7 of 50 times) ( Figure ).

When the chatbots acted as the reviewer, chatbot 1 ranked its own version most favorably, ranking the resident’s version first only 1 of 10 times, the senior author’s version first 3 of 10 times, its own version first 5 of 10 times, and the chatbot 2 version first 1 of 10 times. Chatbot 1 ranked the resident’s version last 3 of 10 times, the senior author’s version last 2 of 10 times, its own version last 2 of 10 times, and the chatbot 2 version last 3 of 10 times. Contrastingly, chatbot 2 was more critical of its own abstracts. Chatbot 2 ranked the resident’s version first 2 of 10 times and the senior author’s version first 2 of 10 times, but it ranked the chatbot 1 version first 6 of 10 times and its own version first 0 of 10 times. Chatbot 2 never ranked chatbot 1 last and ranked itself last 4 of 10 times, the resident last 3 of 10 times, and senior author last 3 of 10 times.

When the frequency of ranks between surgeon-reviewer and chatbot-reviewer was compared, there was no statistical difference in the frequency that the resident or senior author’s abstracts were ranked; however, there was a statistical difference in how the chatbot 1 version and chatbot 2 version were ranked ( Table 1 ). Both the surgeon-reviewers and chatbot-reviewers ranked the resident and senior author’s abstracts similarly, but they ranked chatbot 1 and chatbot 2 abstracts significantly differently. Surgeon-reviewers ranked chatbot 1 abstracts last frequently, while chatbot-reviewers did not. Surgeon-reviewers ranked chatbot 2 abstracts first frequently, while chatbot-reviewers ranked it worse.

There was no statistical difference in the median (IQR) 10-point scores of the resident (7.0 [6.0-8.0]), senior author (7.0 [6.0-8.0]), chatbot 1 (7.0 [6.0-8.0]), or chatbot 2 (7.0 [6.0-8.0]) ( P  = .61). Again, on the 20-point scale, the surgeon-reviewers did not prefer the resident abstracts (median [IQR] score, 14.0 [12.0-17.0]) or senior author’s abstracts (median [IQR] score, 15.0 [13.0-17.0]) over the chatbot 1 (median [IQR] score, 14.0 [12.0-.16.0]) and chatbot 2 versions (median [IQR] score, 14.0 [13.0-16.0]) ( P  = .50). The reviewers’ median (IQR) rank did not differ significantly between abstract versions written by residents (3.0 [1.0-4.0]) or senior authors (2.0 [1.0-4.0]) and abstract versions generated by chatbot 1 (3.0 [2.0-4.0]) or chatbot 2 (2.0 [1.0-3.0]) ( P  = .14) ( Table 2 ). When only comparing the reviews of chatbot 1 and chatbot 2, there was no statistical difference in the 10-point or 20-point scores, but the surgeon-reviewers statistically ranked chatbot 2 better (median [IQR] rank for chatbot 1, 3.0 [2.0-4.0] vs chatbot 2, 2.0 [1.0-3.0]; P  = .02) ( Table 3 ).

When comparing the surgeon-reviewers with chatbot 1 as a reviewer, there was no difference in their 10-point scores, 20-point scores, or ranks of any abstract version. Contrastingly, when comparing the surgeon-reviewers with chatbot 2 as a reviewer, there was a statistical difference in median grades and ranks. Particularly on the 20-point scale, chatbot 2 graded higher than the surgeon-grader for the resident’s abstract version (median [IQR] grade, 14.0 [12.0-17.0] vs 16.9 [16.0-17.5]; P  = .02), the senior author’s abstract version (median [IQR] grade, 15.0 [13.0-17.0] vs 17.0 [16.5-18.0]; P  = .03), the chatbot 1 abstract version (median [IQR] grade, 14.0 [12.0-16.0] vs 17.8 [17.5-18.5]; P  = .002), and the chatbot 2 abstract version (median [IQR] grade, 14.0 [13.0-16.0] vs 16.8 [14.5-18.0]; P  = .04). When the reviews by chatbot 1 and chatbot 2 were compared, again chatbot 2 gave higher median (IQR) grades for all 4 abstract versions on the 20-point scale (resident, 13.5 [13.0-15.0] vs 16.9 [16.0-17.5]; P  = .003; senior author, 13.5 [13.0-15.5] vs 17.0 [16.5-18.0]; P  = .004; chatbot 1, 14.5 [13.0-15.0] vs 17.8 [17.5-18.5]; P  = .003; chatbot 2, 14.0 [13.0-15.0] vs 16.8 [14.5-18.0]; P  = .01). See Table 4 for full analysis.

The first aim of this cross-sectional study was to evaluate if chatbots could generate scientific abstracts as well as a research resident or senior author. Based on 10- and 20-point scales, the abstracts were not differentiable. When force ranked, the chatbot 2 version was ranked first most frequently and the chatbot 1 version was ranked last most frequently. The second goal of this study was to assess how similarly chatbot- and surgeon-reviewers could grade abstracts. Chatbot 1 abstract grades were comparable to the surgeon-reviewers’ grades. However, chatbot 2 graded more favorably than the surgeon-reviewers and chatbot 1. Further observations were that the chatbots consistently utilized the provided results and did not hallucinate new data.

Although editors have worked quickly to regulate the implementation of AI in scientific writing, if it is permitted at all, 14 AI continues to permeate all fields of medicine, academia, and research. 1 , 4 The goal of this study was to evaluate if chatbots could generate and grade medical research abstracts. We found that, when trained using real-world data, chatbots could generate medical research abstracts in a manner that was not able to be differentiated from a human researcher. This is a promising and exciting observation, but further exploration should elucidate the ability of chatbots to consistently grade abstracts, given that the ability varied between chatbots 1 and 2 in our study. There are a variety of rubrics and scoring systems utilized in consideration for national meetings, but our findings indicate that a greater range point-system with defined categories is helpful to discern abstract quality. Abstract grading and consideration are time consuming, but the chatbots showed the potential to expedite this process and could help narrow down the number of abstracts human-reviewers need to read. Our group continues to explore the capability of chatbots as an abstract grader by more extensively training the AI model.

Despite successful implementation of AI in numerous areas of academia, like all new technologies, there is hesitancy to change. 15 Chatbots gather information from unknown sources that cannot be directly cited, leading to controversy over plagiarism and copyright infringement. 4 , 5 To combat this ethical dilemma, some investigators have asked chatbots to provide a list of references, 4 , 13 but when cross-checked, the sources chatbots provided were sometimes falsified. 10 , 11 In the medical field, where patient privacy is extremely important, there is a particular worry about the security of patient information shared with chatbots. 1 , 4 Detractors have labeled chatbots a “stochastic parrot” 1 that “threatens the trajectory” 13 of modern medicine and scientific research. Some believe chatbots will stifle creativity, replace the learned ability of students to write papers, and degrade the sense of academic integrity. 14 , 15 , 76 The counterargument is that learners still develop these writing skills, but in a nontraditional way, by editing chatbot output. 15

Arguably, the most pertinent debate against chatbots is the spread of misinformation. 4 , 10 The hallucinations 18 produced by chatbots may present as fake statistics 77 or inaccurate answers to medical questions. Emile et al 78 assessed a chatbot’s ability to answer common questions about colon cancer, and Samaan et al 79 reviewed the accuracy of a chatbot’s answers regarding bariatric surgery. Both found that the responses were mostly accurate, but there were certainly incorrect answers as well. 78 , 79 Patients using chatbots may not be able to discern fact from fiction, so physicians, whether they support AI or not, should be invested in how their patients are using it. 4 , 15 , 18

Despite these concerns, chatbots have potential in the medical community, including the potential to boost productivity in scientific writing. Chatbots can save researchers time by formatting papers specific to a journal, 1 , 4 running statistics, 18 and accelerating the publishing process, which alleviates pressure on surgeon-scientists. 4 , 6 , 10 , 14 Chatbots can also be leveraged to reduce effort spent preparing a manuscript or grant by editing preexisting text, enhancing readability, and decreasing the number of rounds of feedback between authors. 4 , 10 By increasing efficiency, some believe that chatbots can provide time to devote to more valuable pursuits. 1 , 2 , 4 The ultimate goal of medical research is to advance knowledge and improve health for patients, so if we can employ AI 1 , 4 to perform the routine tasks of research, we can spend more time on the creative aspects, complex questions, and critical thinking involved in research.

Prior studies have investigated the ability of chatbots to regenerate available medical research abstracts. Gao et al 77 provided a chatbot with the title and journal name of previously published abstracts, while Levin et al 23 provided the title and results section and asked it to regenerate the text. Gao et al 77 found that human-reviewers correctly identified 68% of the chatbot-written abstracts and 86% of the human-written abstracts, but the chatbot versions were noted to be vague, making it easier to correctly distinguish them. Levin et al 23 showed that AI-generated versions had fewer grammatical errors and more unique words than the scientist-written version, making these more difficult to distinguish. 24 , 25

This study stands apart from prior work on AI-writing because the chatbots were provided with more than just a title and journal name. 77 By training chatbots to generate text in our group’s writing style and inputting background, previously published studies, and statistically analyzed data for each abstract, we combatted the tendency for chatbots to hallucinate results. We suspect that as chatbots become more sophisticated, the potential to generate abstracts may surpass the ability of some researchers and may expand to generating full manuscripts.

One of the interesting observations we encountered while working with the chatbots was the variation between the chatbots 1 and 2. Both chatbot 1 and chatbot 2 were trained with data extending until September 2021, but chatbot 2 is considered the more advanced version 80 and in our experience, had more independent thinking. 81 When asking the chatbots to generate text, we used the same online session to provide consistency. Chatbot 1 was compliant and completed the tasks without needing redirection, but chatbot 2 had difficulty complying, required restarting new sessions, retraining each one, and several reminders of the prompt to finish writing all 10 abstracts. Although we intended to train the chatbots on more than 10 abstracts, often after the fifth abstract, chatbot 2 pushed back, stating that it did not need more abstracts to learn the writing style. We proceeded, however, in training the chatbots with 10 abstracts. Despite chatbot 2 being less compliant, blinded surgeons agreed that the chatbot 2 abstract versions were better and more consistent than the chatbot 1 versions. The chatbots followed directions on grading more easily, suggesting future promise in saving researchers and editors’ time.

Both advocates and skeptics mostly agree that chatbots will not replace surgeons as primary decision makers in the near future. 4 , 6 , 17 , 21 AI has the potential to complement patient-clinician interactions and assist in medical research, but it will be difficult for AI to replace a surgeon’s judgement. 6 , 17 , 21 Chatbots can serve as a helpful ally in medical abstract generating and grading, but at this point in its evolution, AI cannot perform independently. In the meantime, our goal is to leverage AI for the function of better research and ultimately better patient care. 4 , 14 AI is permeating all facets of medicine, and as clinicians, we need to decide the best approach to incorporate it into our research and clinical space.

The primary limitation of this study was the small sample size of abstracts and reviewers. To combat this limitation, we intentionally chose surgeons who had extensive experience and represented different practice models and international backgrounds. Furthermore, this work is based on abdominal wall reconstruction abstracts and thus may not translate to other fields of medicine. There are also limitations of chatbots. The chatbots have a knowledge cutoff in September 2021 and do not have the ability to browse the internet for more recent context. Chatbots are dependent on the data and training they received, which could result in bias that they learned. 3 , 82 The chatbots additionally have a token cutoff, or character limit, which may inhibit the quantity of training or prompting the model can learn at a time. 17

The findings of this cross-sectional study suggest that a chatbot can generate quality medical research abstracts when the user spends the time to train it, feed it background information, and supply it with analyzed data. The chatbots in this study also demonstrated the ability to grade abstracts, with chatbot 2 being less stringent than chatbot 1. The findings of this study serve as an example of successful and safe implementation of AI in scientific writing, which we hope is considered as editors and publishers continue to determine the regulation and acceptable role of AI.

Accepted for Publication: June 4, 2024.

Published: August 2, 2024. doi:10.1001/jamanetworkopen.2024.25373

Open Access: This is an open access article distributed under the terms of the CC-BY License . © 2024 Holland AM et al. JAMA Network Open .

Corresponding Author: B. Todd Heniford, MD, Division of Gastrointestinal and Minimally Invasive Surgery, Department of Surgery, Atrium Health Carolinas Medical Center, 1025 Morehead Medical Dr, Ste 300, Charlotte, NC 28204 ( [email protected] ).

Author Contributions: Dr Heniford had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Holland, Smart, Ayuso, Scarola, Janis, Fischer, Heniford.

Acquisition, analysis, or interpretation of data: Holland, Lorenz, Cavanagh, Smart, Ayuso, Scarola, Kercher, Jorgensen, Heniford.

Drafting of the manuscript: Holland, Ayuso, Scarola, Kercher, Janis, Fischer.

Critical review of the manuscript for important intellectual content: Holland, Lorenz, Cavanagh, Smart, Ayuso, Scarola, Kercher, Jorgensen, Janis, Heniford.

Statistical analysis: Holland, Scarola.

Administrative, technical, or material support: Lorenz, Cavanagh, Ayuso, Scarola, Kercher, Janis, Heniford.

Supervision: Smart, Ayuso, Scarola, Kercher, Janis, Fischer, Heniford.

Conflict of Interest Disclosures: Dr Smart reported receiving personal fees from Medtronic and Gore outside the submitted work. Dr Heniford reported receiving grants from Gore outside the submitted work. No other disclosures were reported.

Funding/Support: This study was funded by Atrium Health and Carolinas Laparoscopic and Advanced Surgery Program.

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: Neither artificial intelligence nor large language models were employed in the writing of this manuscript.

Data Sharing Statement: See Supplement 2 .

• Register for email alerts with links to free full-text articles
• Access PDFs of free articles
• Manage your interests
• Save searches and receive search alerts