1

Privacy protection in clinical images

(1) Patients’ faces, names, medical record numbers, admission numbers, examination numbers, hospital information, and any other personally identifiable information must be completely masked or de-identified; and (2) De-identification must not affect the interpretation of lesions, anatomical structures, or key objects of observation.

Example:

https://www.wjgnet.com/1948-9366/full/v18/i5/117395-g003.htm

2

Disclosure of AI-generated or AI-assisted figures

(1) AI may be used for mechanism figures, flowcharts, graphical abstracts, and illustrations, but must not be used for Western blots, microscope images, clinical photos, CT, MRI, pathology images, ultrasound images, endoscopy images, or statistical graphs representing real research data; (2) If authors use AI-generated or AI-assisted schematic diagrams, mechanism figures, flowcharts, illustrations, or graphical abstracts, the figure legend must disclose the AI tool name, version, date of use, and main purpose of use; (3) AI-generated or AI-assisted figures must not fabricate experimental results, clinical images, microscopic images, statistical graphs, or any content that could be mistaken for real research data; and (4) Authors are responsible for the accuracy, copyright compliance, and academic integrity of any AI-assisted figures.

3

Use, adaptation, or redrawing of third-party images

(1) Unauthorized use of images is prohibited; (2) If authors directly use, adapt, or redraw third-party images, including images from published medical articles, BioRender, Servier Medical Art, Adobe Stock, Nature Image Library, or similar sources, prior permission must be obtained where required, and the original source, copyright holder, and permission or license information must be stated in the figure legend; and (3) If images are obtained from open-license resources, the license type, source, and conditions of use must be clearly stated.

Example:

https://www.wjgnet.com/1007-9327/full/v32/i5/113592-g003.htm

1

File naming

Figure files should be named according to the format “manuscript number-figure number”. For example, Figure 1 should be named “manuscript number-g001”, Figure 2 as “manuscript number-g002”, and so on.

Example:

https://www.wjgnet.com/2219-2808/figures/v15/i2/113603.htm

2

File format

Figures should be submitted in lossless PNG format. Screenshots of PowerPoint, Word, or Excel pages are not recommended as final publication figures.

3

Resolution

General figures should be 300-500 dpi. Electron microscopy images, high-magnification pathology images, and multi-omics figures should be 500-800 dpi. Figures must meet the required resolution at the final publication size. Authors should not artificially enlarge low-resolution images to achieve a nominal dpi value.

4

Clarity

(1) Figures should be free from blurring, defocusing, mosaic artifacts, or compression distortion; (2) Key objects, such as lesions, cells, bands, and anatomical structures, should have clear edges without obvious blurring; and (3) All panels within the same manuscript should have generally consistent clarity, and obviously blurred panels should be avoided.

5

Figure colors

(1) Figures should use RGB color mode. Colors should be accurate and natural, and excessive color adjustment must not affect the interpretation of data or disease features; (2) Recommended color combinations include blue vs orange, black vs orange, black vs white, and purple vs yellow; and (3) Red vs green should be avoided, especially in bar charts, heatmaps, volcano plots, and UMAP plots, to improve color-blind accessibility.

6

Figure size

The width of a single panel should be 30-168 mm. For composite figures, the maximum width should be 168 mm and the maximum height should be 258 mm.

7

File size

Each figure file should generally not exceed 20 MB. If compression is needed, it must be performed reasonably without compromising publication quality or data authenticity.

8

Panel labels

(1) Panel labels should use uppercase English letters, such as A, B, C, etc.; (2) The font should be bold Tahoma; (3) The font size should be 10 pt; (4) The label color should be black or white and must contrast clearly with the background; and (5) For photographic images, panel labels should be placed in the upper-left corner inside the image, approximately 2 mm from the image edge; for schematic diagrams, statistical graphs, and other non-photographic figures, panel labels should be aligned with the left and upper edges of the image.

Example:

https://www.wjgnet.com/1007-9327/full/v32/i6/114600-g004.htm

9

Panel arrangement and spacing

(1) The spacing between photographic panels should be approximately 1 mm; (2) The spacing between schematic or statistical panels should be 3-5 mm; and (3) Panels should be arranged in alphabetical order from A to Z, from left to right and from top to bottom.

Example:

https://www.wjgnet.com/1007-9327/full/v32/i8/114268-g002.htm

10

Text within figures

(1) Text in figures should comply with the journal’s editorial style, including capitalization of the first word, one space between numbers and units, bold titles, italic lowercase n for sample size, and correct formatting of proper nouns; (2) Regular English and Greek letters should use Tahoma font; (3) The regular font size should be 7 pt. If hierarchy needs to be distinguished, the font size may be adjusted between 5 and 10 pt; and (4) The recommended line spacing is 10 pt.

Example:

https://www.wjgnet.com/1007-9327/full/v32/i9/116223-g001.htm

11

Arrows, triangles, and other markers

(1) Preferred colors include orange (R 255, G 102, B 0), blue (R 0, G 153, B 255), white, or black, with clear contrast against the background; (2) Arrow thickness should be approximately 0.1-0.5 mm; (3) If arrows of different colors, styles, or thicknesses are used in the same figure, their meanings must be explained separately in the figure legend; and (4) If arrows have the same meaning, a consistent arrow style should be used throughout the manuscript.

Example:

https://www.wjgnet.com/2220-3230/full/v16/i2/116999-g001.htm

12

Figure legends

(1) Each panel label, such as A, B, or C, must correspond exactly to the description in the figure legend; (2) All abbreviations used in the figure must be defined in the legend; (3) Statistical symbols, such as *, **, ***, #, ##, and ###, should be converted to letter annotations and explained clearly in the figure legend; and (4) Figure legends should explain key markers, arrows, scale bars, staining methods, magnification, and statistical meanings.

Example:

https://www.wjgnet.com/1007-9327/full/v32/i14/114331-g001.htm

13

Composite figures

(1) Line thickness should be consistent throughout the figure; (2) Panels of the same type should have similar or identical dimensions; (3) Alignment principles should be followed, for example, the x-axes of two adjacent bar charts should be aligned; (4) Large blank spaces within figures should be avoided, and the overall layout should be neat, compact, and visually clear; and (5) Panels from different sources, experimental conditions, or time points must not be assembled in a misleading manner.

Example:

https://www.wjgnet.com/1007-9327/full/v32/i9/114207-g001.htm

1

Grayscale images, such as CT, MRI, electron microscopy, and light microscopy images

(1) All microscopy and electron microscopy images must include a scale bar. The scale bar should preferably be white or black, approximately 0.2-0.4 mm thick, and placed in the lower-right corner. The scale bar must be indicated either within the image or in the figure legend. Common scale bars include 100 μm, 50 μm, or 10 μm; (2) Scale bars, magnification, and text within the figure must be clearly readable; (3) Grayscale levels should be complete, without obvious scanning artifacts, overexposure, underexposure, or abnormal compression marks; and (4) Lesions or key structures should be clearly identifiable.

Example:

https://www.wjgnet.com/2218-5836/full/v17/i6/117244-g002.htm

2

Color clinical images, such as pathology, endoscopy, surgery, and specimen images

(1) All pathology images must include a scale bar. The scale bar should preferably be white or black, approximately 0.2-0.4 mm thick, and placed in the lower-right corner. The scale bar must be indicated either within the image or in the figure legend. Common scale bars include 100 μm, 50 μm, or 10 μm; (2) Patient privacy information must be completely masked or de-identified; (3) Scale bars and marker lines should be clean and clear; (4) Lesion areas must not be obscured by reflections, covering objects, or markers; and (5) Colors should be as close as possible to the actual observation and must not be excessively adjusted.

Example:

https://www.wjgnet.com/2307-8960/full/v14/i17/121921-g001.htm

3

Line drawings, such as mechanism figures, flowcharts, and PRISMA flow diagrams

(1) Approximately 2 mm of blank space should be kept around the text inside text boxes; (2) Text should be center-aligned or left-aligned, and formatting should be consistent throughout the figure; (3) Lines, arrows, and graphic elements should have consistent thickness; and (4) PRISMA flow diagrams should clearly show the stages of Identification, Screening, Eligibility, and Included. Stage labels are recommended to use an orange background fill (R 255, G 102, B 0) with bold white text.

Example:

https://www.wjgnet.com/2220-3249/full/v15/i2/118445-g001.htm

4

Electrophoresis, Western blot, and PCR gel images

(1) Band boundaries should be clear and sharp, without obvious smearing; (2) Group names, sample order, and bands must correspond exactly; (3) Unexplained spots, splicing traces, or abnormal backgrounds must not be present; (4) For Western blots, the protein name, loading control, molecular weight, exposure condition, and replicate number must be stated; (5) Individual bands must not be selectively adjusted, deleted, cropped, or spliced without explanation; and (6) If cropping or splicing has been performed, this must be explained in the figure legend, and verifiable original images, uncropped full-membrane images, full-gel images, or complete scanned images must be retained.

Example:

https://www.wjgnet.com/2220-315X/full/v16/i2/118250-g001.htm

5

Statistical graphs, such as bar charts, line graphs, and scatter plots

(1) Axes should be black, with a line thickness of approximately 0.1 mm; (2) The distance between axis labels, values, and tick marks should be approximately 1 mm; (3) Bar charts should avoid pattern fills, and different grayscale levels or clearly distinguishable colors are preferred; (4) Error bars should be clearly outlined; (5) Scatter points should not merge into indistinct clusters, and curve edges should be smooth; (6) Statistical graphs should provide the statistical test method, sample size, mean, median, SD, SEM, 95%CI, IQR, and other necessary information, and the meaning of error bars and significance annotations must be explained in the figure legend; and (7) Bar charts should avoid hiding raw data where possible. Scatter plots, box plots, and violin plots are recommended.

Example:

https://www.wjgnet.com/1948-5204/full/v18/i5/116568-g001.htm

6

Multi-omics figures

Multi-omics figures, including heatmaps, volcano plots, PCA, UMAP, and tSNE plots, should use consistent fonts and colors, and legends must be complete. Color gradients and grouping information should be clearly identifiable.

Example:

https://www.wjgnet.com/1007-9327/full/v32/i17/116386-g003.htm

7

Graphical Abstract

A graphical abstract should summarize the key findings of the study. The recommended maximum width is 168 mm and the maximum height is 258 mm. The text should not exceed 50 words. Complex backgrounds should be avoided, and references or statistical results should not be included.

Example:

https://www.wjgnet.com/1948-5182/full/v18/i6/118607-g001.htm

1

Consistency with original images

Authors must ensure consistency among the original images, published figures, and original data. Authors must not replace experimental replicate images, Western blot bands, microscopic fields, or animal images unless this is clearly explained in the figure legend and supported by verifiable original evidence.

2

Consistency with data

Figures must be consistent with tables, raw data, supplementary materials, and statistical analyses. For example, heatmaps must not contain data absent from the main text, and scatter plots must not be inconsistent with the original Excel data.

3

Permitted image processing

Any image enhancement must be applied to the entire image. Uniform adjustment of brightness, contrast, levels, gamma, white balance, rotation, cropping, slight sharpening, and slight noise reduction is permitted. Cropping should be explained in the figure legend when necessary.

4

Prohibited image processing

Figures must not be selectively altered in specific areas. Authors must not delete experimental data, add experimental data, duplicate local regions, over-beautify images, use Clone Stamp, Heal Brush, Content-Aware Fill, AI image repair, remove backgrounds, remove spots, remove bands, or remove cells.

1

Does the file name follow the format “manuscript number-g001/g002…”?

2

Is the figure submitted in lossless PNG format?

3

Does the resolution meet publication requirements?

4

Is the figure clear and free from blurring, defocusing, mosaic artifacts, or compression distortion?

5

Have all clinical images been properly de-identified to protect patient privacy?

6

Have all reused, adapted, or redrawn third-party images received appropriate permission and source attribution?

7

Have all AI-generated or AI-assisted figures disclosed the tool name, version, date of use, and purpose in the figure legend?

8

Are panel labels A/B/C standardized, clear, and in the correct order?

9

Do the font, font size, units, and abbreviations within the figure comply with journal style?

10

Are arrows, scale bars, markers, and symbols clear and explained in the figure legend?

11

Is the composite figure aligned, compact, visually clear, and consistent in panel clarity?

12

Does the figure legend correspond to all panels and explain abbreviations and statistical symbols?

13

For Western blot/PCR/gel images, have the original full-membrane, full-gel, or complete scanned images been retained for verification?

14

Do statistical graphs explain sample size, error bar meaning, and statistical methods?

15

Is the figure free from misleading splicing, excessive processing, or inconsistency with the underlying data?

1

Figure submission by authors

Authors must submit figures that are accurate in content, properly formatted, clear, complete, and compliant with ethical and copyright requirements.

2

Image integrity screening

The editorial office may use tools such as ImageTwin, Proofig, Forensically, ImageJ, or ORI Image Forensics to detect duplicated images, mirroring, rotation, local duplication, abnormal splicing, excessive processing, or traces of AI generation. This process is considered Image Integrity Screening and serves only as a risk alert and basis for editorial review. It must not be used as the sole basis for determining academic misconduct.

3

Editorial review and grading

Editors review figures according to publication standards, manuscript content, and figure legends. The grading and handling measures are as follows: (1) Grade A: Meets publication standards and requires no modification; (2) Grade B: Minor technical issues that may be corrected by the editorial office; (3) Grade C: Publication-quality problems requiring author revision; (4) Grade D: Potential image authenticity or data consistency risk requiring image authenticity verification; and (5) Grade E: Suspected academic misconduct requiring a publication ethics investigation.

4

Return for author revision

Reasons for returning figures for revision include, but are not limited to: (1) Patient privacy information is not masked or is insufficiently masked; (2) Unauthorized images are used, or image sources and permission information are not provided; (3) Permission or licensing documentation is incomplete; (4) AI-generated or AI-assisted figures do not disclose the tool name, version, date of use, and purpose; (5) Figure resolution is insufficient and affects publication quality; (6) Western blot, PCR, or electrophoresis images show suspected abnormal splicing, duplicated bands, or inconsistent backgrounds; (7) Figure legends do not correspond to panels, or key abbreviations, arrows, scale bars, or statistical symbols are not explained; (8) Statistical graphs are inconsistent with the main text, tables, or original data; (9) Text within figures does not comply with editorial style; and (10) Composite figure layout is disorganized and affects reader understanding or publication quality.

5

Temporary suspension of publication

Publication may be temporarily suspended for reasons including, but not limited to: (1) Figure resolution is insufficient and affects final publication quality; (2) Statistical graphs are inconsistent with the main text, tables, or original data and require author explanation or correction; (3) Composite figure layout is disorganized and affects reader understanding or publication quality; and (4) Authors are temporarily unable to provide necessary original images, original data, or permission documentation.

6

Academic integrity investigation

An academic integrity investigation may be initiated for reasons including, but not limited to: (1) Unauthorized use of images, or failure to provide image source and permission information; (2) Undisclosed AI-generated or AI-assisted figures that may affect the assessment of research authenticity; (3) Western blot, PCR, or electrophoresis images showing suspected abnormal splicing, duplicated bands, or inconsistent backgrounds; (4) Statistical graphs that are clearly inconsistent with the main text, tables, or original data; (5) Images showing obvious artifacts, excessive beautification, misleading cropping, or inappropriate image processing; and (6) Authors are unable to provide necessary original images, original data, or permission documentation.

7

Final publication

After author revision, editorial re-checking, and, where necessary, image authenticity verification, figures may be published with the accepted manuscript only when they meet the journal’s requirements for publication quality, ethics, copyright compliance, and academic integrity.

1

Figure retention

(1) Authors must retain original photographs, RAW files, TIFF files, uncompressed PNG files, original microscopy images, original Western blot images, full-membrane images, full-gel images, and complete scanned images; and (2) These files should be retained for at least 10 years, or for a longer period if required by the authors’ institution, funder, or applicable regulations.

2

Original data management

The editorial office may request original images, metadata, acquisition dates, original data tables, experimental records, or permission documentation at submission, peer review, revision, before acceptance, before publication, or after publication when necessary to verify image authenticity, data consistency, and copyright compliance.