Image integrity & digital figures

Policy on the preparation, processing and presentation of scientific images and digital figures in IUMS journals, including what adjustments are acceptable, what is prohibited, and how image-related concerns are handled.

Applies to all figures, images and image-based data
Policy v1.0 – last updated April 2025
Related policies: Research ethics & integrity · Research data & code sharing · Research & publication misconduct · Corrections & retractions

1 Scope & purpose

Why image integrity matters for IUMS journals.

Scientific images – such as microscopy, radiology, pathology, gels, blots and graphical renderings – are central to many articles in IUMS journals. Readers, reviewers and editors rely on these images to interpret and trust the reported findings. Inaccurate or misleading image manipulation can undermine confidence in the entire study and, in serious cases, in the broader literature.

This policy sets out:

  • what kinds of digital adjustments are acceptable;
  • which practices are considered inappropriate or fraudulent;
  • how authors should document image processing;
  • how IUMS journals screen for and respond to image-related concerns.

The policy covers all image types used in submissions, including figures, supplementary images, graphical abstracts and images generated by digital or AI-assisted tools.

2 Core principles of image integrity

Images must faithfully represent the underlying data.

IUMS journals follow three core principles:

  • Fidelity: images must not misrepresent the original data. Adjustments should not add, remove or change features in a way that could alter scientific interpretation.
  • Transparency: relevant image acquisition and processing steps should be described sufficiently so that readers and reviewers understand how images were produced.
  • Accountability: authors are responsible for the integrity of all images in their submissions, even when preparation involves third parties or core facilities.

Minor aesthetic improvements are allowed, but they must not cross the line into fabrication, falsification or deceptive enhancement.

3 Acceptable image adjustments

Adjustments that are allowed when applied carefully and consistently.

The following adjustments are generally acceptable, provided that they are applied to the entire image (or to all images in a panel consistently), do not remove real signals or introduce artefacts, and are reported when relevant:

  • Global brightness and contrast: making the overall image clearer, without saturating or clipping important regions.
  • Global colour balance: adjusting colour channels to improve readability, provided all images in a set are treated in the same way.
  • Linear scaling: applying linear adjustments to intensity or signal ranges, with no selective enhancement of specific areas.
  • Minor cropping: removing extraneous borders or irrelevant regions to focus on the area of interest, as long as cropping does not remove critical features or misrepresent the context.
  • Rotation and alignment: correcting orientation or alignment so that images can be compared more easily.

Where image processing settings are important to interpretation (for example in quantitative fluorescence imaging), authors should provide the main parameters used in the Methods or figure legends.

4 Documenting image acquisition & processing

Basic reporting of how images were generated and adjusted.

For image-heavy articles, the Methods section should briefly describe:

  • the imaging modality and equipment (e.g. microscope type, objective, detector, scanner, imaging software);
  • key acquisition parameters (e.g. exposure times, magnification, voxel size);
  • any major processing steps applied to the raw images (e.g. deconvolution, background subtraction, filtering).

When non-trivial digital processing is applied – for example, 3D reconstructions, projections, segmentation or thresholding – the main steps and software tools used should be described clearly enough for a knowledgeable reader to understand what has been done.

Detailed processing pipelines, macros or scripts can be shared as supplementary material or via repositories, in line with the Research data & code sharing policy.

5 Unacceptable image manipulation

Practices that mislead readers or distort the scientific record.

The following practices are not acceptable and may be treated as research or publication misconduct:

  • Selective enhancement or suppression: altering only specific regions (for example, enhancing some bands in a gel, selectively brightening part of a microscopy image) without applying the same change to the entire image or panel.
  • Adding, removing or moving features: inserting, deleting or repositioning bands, cells, lesions or other visual elements to change the appearance of results.
  • Cloning or duplicating elements: copying parts of an image and pasting them elsewhere to fill gaps, create or remove signals or “clean up” backgrounds.
  • Non-linear manipulation that distorts data: extreme non-linear adjustments that exaggerate or diminish signals in ways that could mislead interpretation.
  • Misleading contrast/brightness: adjusting levels so strongly that bands disappear, background becomes artificially uniform or signal saturation hides real differences.
  • Inconsistent processing within a panel: applying different processing to images that are presented for direct comparison, without clear indication and justification.

Minor cosmetic edits such as removing dust or isolated scanner artefacts are discouraged and, if ever used, must not alter scientific meaning. Systematic removal of noise or artefacts that changes measurable values is not acceptable.

6 Composites, splicing & reuse of images

Combining panels and using images across figures.

Combining multiple images into a single figure (for example, multi-panel composites or spliced gels) is allowed under the following conditions:

  • any boundaries between spliced lanes or fields are clearly indicated (e.g. with thin dividing lines or spacing);
  • panels originate from the same experiment and are appropriate for side-by-side comparison;
  • the figure legend clearly explains how the composite was constructed if it is not obvious from the layout.

Reuse of the same image to represent more than one experimental condition, specimen or time point is not acceptable unless this reuse is explicitly declared and scientifically justified (for example, the same control image used to accompany multiple treatments within one experiment).

Authors should retain original uncropped, unspliced versions of gels and blots and be prepared to provide them if requested by editors or reviewers.

7 Microscopy, radiology, gels & blots

Additional expectations for common scientific image types.

7.1 Microscopy & fluorescence imaging

For microscopy images, authors should:

  • include scale bars on all images where size matters to interpretation;
  • ensure that any pseudo-colouring is explained in the legend;
  • use consistent acquisition and processing settings when presenting images for quantitative comparison;
  • avoid over-smoothing or aggressive filtering that removes real structures.

7.2 Radiology & clinical imaging

For radiology and other clinical imaging:

  • any patient-identifying information embedded in the image (e.g. names, IDs, dates) must be removed in line with the Consent & privacy policy;
  • windowing and level settings should not obscure clinically relevant findings;
  • annotations, arrows and labels must be accurate and clearly visible.

7.3 Gels, blots & similar assays

For gels, blots and similar assays, authors should:

  • avoid duplicating bands or lanes across different figures or experiments;
  • indicate when non-adjacent lanes from the same gel are presented next to each other;
  • describe any background subtraction, normalisation or other processing applied;
  • provide uncropped, original images as supplementary files when required by the journal.

8 Raw data, original files & retention

Keeping original image data available for verification.

Authors are expected to:

  • retain original, unprocessed image files (e.g. instrument-generated formats, uncropped gels) for an appropriate retention period, typically at least several years after publication, in line with institutional policies;
  • be able to supply original image data to editors if requested during peer review or post-publication investigations;
  • ensure that original files are securely stored and backed up in ways that protect participant privacy where applicable.

In some cases, journals may require deposition of image datasets in repositories, particularly for high-throughput imaging studies, in line with the Research data & code sharing policy.

9 Patient images & clinical photographs

Images of identifiable individuals require explicit consent.

Clinical photographs and other images that show recognisable patients or participants are subject to strict ethical and legal requirements. Authors must:

  • obtain explicit, written consent for publication from the patient or their legal guardian, in line with journal and institutional policies;
  • state in the manuscript that such consent was obtained;
  • remove direct identifiers (e.g. names, hospital numbers, exact dates) and consider whether individuals could still be indirectly identifiable;
  • avoid using masking techniques that may be easily reversible or give a false sense of anonymity.

For images where consent cannot be obtained (for example, older archival material), publication will be considered only in exceptional circumstances and must comply with local regulations and the Consent & privacy policy.

10 Editorial screening & investigation of concerns

How IUMS journals check images and respond to potential problems.

IUMS journals may use a combination of visual inspection, expert review and software tools to screen images before and after publication. If potential problems are identified, editors may:

  • request original image files and a detailed explanation of processing steps from the authors;
  • seek independent expert opinion on whether manipulation appears acceptable or problematic;
  • ask authors to supply corrected figures with clearer labelling or more transparent processing.

When concerns arise from readers or reviewers, they will be handled in line with the Research & publication misconduct policy, which may include contacting authors’ institutions or funders if serious issues are suspected.

11 Corrections, retractions & learning from cases

Possible outcomes when image integrity issues are confirmed.

Depending on the nature and severity of the problem, and whether it can be corrected without undermining the main findings, possible editorial actions include:

  • Corrections: replacing a figure, clarifying processing in the legend or adding an explanatory note when the underlying data remain reliable.
  • Expression of concern: an interim notice when serious questions exist but investigations are ongoing or incomplete.
  • Retraction or withdrawal: when image manipulation seriously compromises or invalidates the findings, or when there is evidence of deliberate data fabrication or falsification.

In all cases, IUMS journals aim to follow recognised best practice for corrections and retractions, as described in the Corrections & retractions policy, and to use cases to improve guidance and education for authors and reviewers.

12 Roles & responsibilities

Who is responsible for maintaining image integrity.

Responsibilities are shared, but authors hold primary responsibility for the images they submit:

  • Authors:
    • ensure that all images and figures comply with this policy;
    • retain and provide original image data when requested;
    • describe acquisition and processing methods transparently;
    • avoid delegating image handling to individuals who are not familiar with these requirements.
  • Institutions & laboratories:
    • provide training on good image handling practices;
    • establish internal procedures for secure storage and responsible processing of imaging data.
  • Editors & journals:
    • communicate clear expectations to authors and reviewers;
    • use proportional screening and follow-up mechanisms;
    • apply policies consistently and in line with broader research integrity frameworks.
  • Reviewers and readers:
    • flag possible concerns about image integrity confidentially to editors, without making public allegations.

13 Questions & contact

Getting help with image preparation and integrity issues.

Authors who are unsure whether a particular adjustment or composite layout is acceptable should contact the editorial office of the target IUMS journal before or during submission. Where possible, provide example images and a brief description of the processing steps applied.

General questions about this policy can be directed to:

Policy version: v1.0 – last updated April 2025. This page will be reviewed periodically in light of evolving good practice and technological developments in image analysis.