8.1 Purpose and Scope
- This policy safeguards the accuracy of all visual and audiovisual materials in published content: scientific images, charts, diagrams, artwork, audio, video, animations, and interactive media. It applies to authors, editors, reviewers, and production staff across all journals we publish.
8.2 First Principles
- Show the data, do not reshape the truth. Visual changes must never mislead or alter scientific meaning.
- Transparency by default. Any enhancement or composition must be disclosed in the figure legend or Methods.
- Verifiability. Originals and processing steps must be available for editorial inspection upon request.
8.3 Acceptable Global Adjustments (with disclosure)
- Uniform, non-selective adjustments are allowed if they do not hide, introduce, or exaggerate features:
- Brightness, contrast, and gamma applied to the entire image.
- Uniform white balance, exposure, or color-space conversions.
- Linear cropping that does not remove essential context (e.g., scale bars, controls).
- Disclosure: “Global brightness/contrast adjustments applied; no pixel-level alterations.”
8.4 Prohibited Manipulations
- Selective enhancement, suppression, or removal of features (e.g., erasing bands, adding spots, cloning backgrounds).
- Splicing lanes or fields without clear demarcation lines and labeling.
- Nonlinear intensity adjustments that distort relative differences without disclosure.
- Artificial smoothing/denoising that removes meaningful signal or artifacts critical for interpretation.
- AI-based inpainting or generative fills that fabricate data elements.
- Violations may trigger rejection, correction, expression of concern, or retraction.
8.5 Composites and Splices (Permitted with strict labeling)
- Show borders between panels/lanes/fields with clear separating lines.
- Label each segment (e.g., “lane 1–4 from gel A; lane 5–7 from gel B”).
- Provide acquisition dates/instruments if multiple sessions were combined.
- State any re-ordering of lanes or fields.
8.6 Source Data and Provenance
- For every figure derived from empirical data, authors must retain and, on request, provide:
- Raw, unprocessed image files (native formats, e.g., .lif, .czi, .nd2; DICOM for radiology).
- Acquisition metadata (instrument, objective, numerical aperture, voxel size/pixel scale, exposure, filters).
- Processing logs or stepwise description sufficient to reproduce the final figure.
- Minimum retention: consistent with field norms and legal/ethical requirements.
8.7 Scale, Units, and Calibration
- Include scale bars or explicit magnification and units.
- Do not rely on relative magnification descriptions alone.
- Report calibration methods for imaging systems and measurement devices.
8.8 Color, False Color, and Perceptual Integrity
- Use color maps that preserve ordinal/interval relationships; avoid misleading palettes that create false boundaries.
- Identify false-color or pseudocolor images and specify channel mapping.
- Multi-channel images must state dyes/fluorophores, emission/ excitation settings, and channel order.
8.9 Specific Modalities — Minimum Standards
- Gels/Blots: Show full-length blots or provide uncropped source images as supplements; mark reused lanes; avoid background obliteration.
- Microscopy: Report optical setup, z-stacks, deconvolution, projection type; disclose stitching/tiling.
- Flow Cytometry: Provide gating strategy, compensation, controls, event counts; show axis scales consistently.
- Spectra/Chromatograms: Include axis units, baselines, smoothing parameters; avoid baseline subtraction without disclosure.
- Radiology/Clinical: Anonymize DICOM headers; disclose window/level settings if diagnostically relevant.
- Maps/Geospatial: Name coordinate reference systems, data sources, and spatial resolution; indicate any generalization or masking for privacy.
8.10 Charts and Statistical Graphics
- Axes must have units, scales, and zero baselines where appropriate.
- Avoid truncated axes that distort effect size; if truncation is necessary, indicate clearly.
- Show individual data points or distribution summaries (e.g., box/violin plots) whenever feasible; avoid deceptive 3D embellishments.
- Report sample sizes (n) and measures of dispersion (SD/SE/CI) in legends or captions.
8.11 Figure Assembly and Layout
- Panels labeled consistently (A, B, C…); legends explain each panel.
- Do not duplicate the same image in multiple panels to represent different conditions.
- Avoid excessive compression artifacts; submit publication-quality resolution (typically ≥300 dpi for raster images at final size).
8.12 Audio, Video, and Animation
- State frame rate, resolution, and any edits (speed changes, cuts, stabilization).
- Do not add or remove sounds or visual events that change interpretation.
- Provide closed captions or transcripts when feasible; anonymize voices and faces where required.
8.13 Accessibility of Visuals
- Provide alt text or concise, descriptive captions that convey the scientific message.
- Ensure sufficient contrast ratios for text and critical features.
- Avoid color-only encodings; use patterns or markers to differentiate series.
8.14 Image Integrity Screening (Publisher)
- Editorial staff may use automated tools to identify:
- Duplications, splices, contrast anomalies, and cloning artifacts.
- Inconsistent noise patterns or compression signatures.
- Findings trigger human expert review; automated flags never determine outcomes alone.
8.15 Disclosure Templates (for authors)
- Image Processing Statement: “Only uniform, global adjustments (brightness/contrast) were applied. No selective editing or data removal occurred. Composite boundaries are marked; full-length blots provided in Supplementary Figures S1–S3.”
- Figure Legend Minimums: source of data; acquisition instrument/settings; units/scale; n per group; statistical tests; definitions of symbols and error bars.
8.16 Third-Party and Stock Visuals
- Scientific illustrations or icons must not imply empirical measurements.
- Obtain and document licenses/permissions.
- Acknowledge illustrators/designers; avoid misleading photorealistic renderings for hypothetical mechanisms.
8.17 Privacy and Identifiability
- Remove direct identifiers from clinical images or obtain explicit consent for publication.
- For sensitive locations (e.g., endangered species habitats, refugee settlements), consider obfuscation or coordinate generalization and state the method.
8.18 Non-Compliance and Remedial Actions
- Pre-publication: Request originals, revisions, or withdraw the manuscript.
- Post-publication: Publish corrections, image replacements with clear notes, expressions of concern, or retractions if conclusions are undermined.
- Notify institutions/funders when appropriate; apply submission bans for repeated or egregious violations.