Image Manipulation

1. Purpose and Scope

AMG Transcend Association requires images, figures, and other visual representations of research data to accurately and transparently reflect the underlying evidence.

This policy establishes publisher-level principles concerning the preparation, processing, presentation, assessment, and post-publication handling of image-based research materials across the journals published by AMG Transcend Association.

It applies, as relevant, to authors, reviewers, editors, Editorial Board members, and other participants in the editorial and publication process.

Individual journals may establish more detailed requirements appropriate to their disciplines, imaging methods, article types, and editorial procedures.

2. Fundamental Principle of Image Integrity

Images and other visual representations of research data must not be fabricated, falsified, selectively altered, duplicated, obscured, enhanced, reconstructed, or otherwise manipulated in a manner that misrepresents the underlying evidence or could materially mislead scientific interpretation.

Image preparation and processing must preserve the scientific meaning, provenance, and integrity of the original data.

Authors remain responsible for the integrity and accurate presentation of all images included in their manuscripts, including images prepared or processed by collaborators, technicians, core facilities, service providers, or automated systems.

3. Legitimate Image Processing

Scientifically legitimate image processing is permitted where it is appropriate to the research method and does not create a false or misleading representation of the underlying data.

Permissible processing may include, where scientifically appropriate:

• global adjustments to brightness, contrast, or color balance;
• cropping that does not remove relevant scientific context;
• appropriate resizing or formatting;
• scientifically justified pseudocoloring or color mapping;
• processing or filtering that forms a legitimate part of the analytical method;
• annotations, labels, scale bars, or other explanatory elements.

Adjustments intended only to improve presentation should ordinarily be applied consistently to the complete image.

Processing must not selectively eliminate, introduce, conceal, exaggerate, or reposition scientifically relevant features.

Where processing materially contributes to the analysis or interpretation of the data, the relevant methods must be described with sufficient transparency for scientific evaluation.

4. Unacceptable Image Manipulation

Unacceptable practices include, but are not limited to:

• adding, removing, concealing, moving, or duplicating scientifically relevant image features;
• selective enhancement or suppression of particular regions, signals, bands, cells, structures, or other data elements in a misleading manner;
• inappropriate local adjustment of brightness, contrast, color, or other image properties;
• duplication or partial duplication of an image to represent different samples, experiments, conditions, time points, or results;
• flipping, rotating, resizing, cropping, or otherwise modifying an image for the purpose of concealing inappropriate duplication;
• inappropriate splicing or assembly that misrepresents separate images as a single continuous observation or experiment;
• concealment of boundaries between images originating from different experiments, acquisitions, or fields where that distinction is material;
• manipulation of controls, backgrounds, baselines, or comparative images in a manner that distorts the experimental context;
• generation, reconstruction, or alteration of image-based research evidence in a manner that falsely represents generated information as an original observation;
• other modifications that materially misrepresent the provenance, content, or scientific meaning of the visual evidence.

Intentional deception is not required for an image to be scientifically unreliable. Editorial assessment should distinguish, however, between honest error, inappropriate processing, inadequate reporting, and possible deliberate falsification.

5. Composite Images, Splicing, and Figure Assembly

Composite and multi-panel figures are permitted where their structure and provenance are transparent.

Where images from different acquisitions, experiments, fields, samples, or conditions are combined:

• individual components must be distinguishable where necessary for interpretation;
• relevant boundaries or discontinuities must not be concealed;
• assembly must not create the false appearance of a continuous observation or experiment;
• reused image elements must be identified where legitimate reuse is scientifically justified;
• figure legends or methods must provide sufficient information to understand the relevant assembly or processing.

Undisclosed duplication, inappropriate reuse, or misleading assembly of visual elements is not acceptable.

6. Original Images and Supporting Information

Authors must retain original or source image files and relevant supporting information in accordance with applicable institutional, legal, regulatory, contractual, funder, and disciplinary requirements.

Where reasonably necessary to assess a manuscript or investigate a credible integrity concern, editors may request:

• original or unprocessed image files;
• relevant metadata;
• processed image versions;
• acquisition information;
• processing or analytical documentation;
• information concerning software, tools, settings, or workflows;
• other materials necessary to understand the provenance and preparation of the image.

Requests must be proportionate to the issue under assessment and remain subject to legitimate privacy, confidentiality, legal, ethical, intellectual-property, security, and data-protection restrictions.

Failure or inability to provide requested original materials may affect the journal’s ability to verify the reliability of the work but does not by itself establish misconduct or image manipulation.

7. Documentation and Transparency

Image acquisition and processing must be reported with sufficient detail for appropriate scientific evaluation where those procedures materially affect the findings or their interpretation.

Authors should disclose, where relevant:

• the imaging method or instrument;
• important acquisition conditions or settings;
• material image-processing procedures;
• analytical or computational transformations;
• software or tools used where relevant to interpretation or reproducibility;
• composite assembly, splicing, or material restructuring of figures;
• other processing necessary for readers and reviewers to understand how the reported visual evidence was produced.

Routine presentation adjustments need not be described exhaustively where they do not materially affect scientific interpretation, unless required by the applicable journal.

8. Artificial Intelligence and Automated Image Processing

Artificial intelligence, machine-learning systems, automated processing, and computational imaging methods may form part of legitimate scientific methodology.

Where such tools materially contribute to image acquisition, processing, analysis, segmentation, reconstruction, classification, enhancement, or generation, their use must be scientifically justified and reported with sufficient transparency for evaluation.

AI or automated tools must not be used to:

• fabricate experimental or observational images;
• introduce features not supported by the underlying evidence;
• remove or conceal scientifically relevant features without legitimate methodological justification;
• reconstruct missing information and present it as directly observed empirical evidence;
• disguise duplication or other inappropriate manipulation;
• generate misleading representations of research findings.

AI-generated illustrations or conceptual images that do not represent empirical observations must be clearly distinguishable from research data where there is a reasonable risk of confusion.

Original underlying data should be preserved where AI or computational processing produces derived visual outputs.

Applicable requirements of the AI and Generative Tools policy remain in force.

9. Images Involving Human Participants

Images involving identifiable human participants must comply with applicable ethical, privacy, confidentiality, informed-consent, and data-protection requirements.

Image alteration, masking, cropping, or blurring does not automatically establish that an individual is no longer identifiable.

Where consent for publication of identifiable visual material is required, the applicable requirements of the Human and Animal Research Ethics and Informed Consent and Clinical Trials policies must be followed.

10. Responsibilities of Authors

Authors are responsible for ensuring that:

• images accurately represent the underlying evidence;
• prohibited image manipulation has not occurred;
• legitimate processing is scientifically justified;
• material processing is appropriately documented and reported;
• figure legends and labels accurately describe the visual content;
• relevant original or supporting materials are retained appropriately;
• third-party image rights, permissions, confidentiality, and ethical requirements are respected;
• AI-assisted or automated processing is disclosed where required.

Authors must cooperate with legitimate editorial requests concerning image integrity.

11. Responsibilities of Reviewers

Reviewers should assess image-based evidence within the reasonable limits of their expertise and the purposes of peer review.

Potential concerns such as apparent duplication, inconsistent image presentation, unexplained splicing, inappropriate processing, or other suspected integrity problems should be communicated confidentially to the responsible editor.

Reviewers are not expected to conduct forensic image investigations or independently establish research misconduct.

Confidential manuscript images must not be uploaded to external image-analysis, artificial-intelligence, or other third-party systems where doing so would violate applicable confidentiality, privacy, intellectual-property, or data-protection requirements.

12. Editorial Assessment

Editors may examine image-based materials and request clarification or supporting information where reasonably warranted by scientific, editorial, ethical, or integrity concerns.

Assessment may include:

• examination of original or source images;
• comparison between original and processed files;
• review of relevant metadata or processing documentation;
• additional peer review;
• consultation with an appropriately qualified independent expert;
• assessment of related figures or publications where relevant;
• referral to another applicable publication-ethics procedure.

Automated or forensic tools may support assessment where appropriate, but tool-generated scores, classifications, or detections must not by themselves establish that inappropriate manipulation or misconduct has occurred.

Editorial conclusions must be based on the available evidence and appropriate scientific judgment.

13. Suspected Image Manipulation

Potential image-integrity concerns must be assessed fairly, confidentially, and proportionately.

Authors should normally be given an appropriate opportunity to explain the issue and provide relevant source materials or documentation before an adverse editorial conclusion is reached.

The journal must distinguish, where reasonably possible, between:

• legitimate scientific processing;
• inadequate or unclear reporting;
• preparation or presentation errors;
• inappropriate image manipulation;
• serious falsification or fabrication requiring possible external investigation.

Where formal investigative authority is required, the journal or publisher may refer the matter to an appropriate institution, research organization, ethics body, funder, regulator, or other competent authority.

The journal remains responsible for determining the appropriate editorial status of the manuscript or publication within its scholarly record.

14. Editorial and Post-Publication Action

Before publication, image-integrity concerns may result, as appropriate, in:

• clarification;
• provision of original or supporting materials;
• correction or replacement of a figure;
• revision of methods, legends, or disclosures;
• additional scientific or expert assessment;
• rejection where the reliability or integrity of the work cannot reasonably be established;
• institutional or other external referral where warranted.

After publication, appropriate action may include:

• no further action where the concern is not substantiated;
• clarification;
• Correction;
• Expression of Concern;
• retraction;
• external referral;
• another proportionate action necessary to protect the scholarly record.

Post-publication actions must follow the Publication Ethics and Research Integrity and Corrections, Retractions and Expressions of Concern policies.

15. Relationship with Other Policies

This policy should be read together with applicable AMG Transcend Association and journal policies concerning:

• Data Integrity and Reproducibility
• Publication Ethics and Research Integrity
• AI and Generative Tools
• Research Ethics
• Human and Animal Research Ethics
• Informed Consent and Clinical Trials
• Peer Review Principles
• Authorship and Contributor Responsibilities
• Corrections, Retractions and Expressions of Concern
• Complaints and Appeals

Individual journals may establish additional image-integrity, figure-preparation, technical, methodological, or documentation requirements appropriate to their disciplines and published content.