Integrity Assessment: "Altered microstructural pattern of the cortex and basal forebrain cholinergic system in Wilson's disease: an automated fiber quantification tractography study" (Brain Imaging and Behavior, 2023)

Academic fraud report · Geng Detector

Summary

Verdict: Questionable (yellow flag), with concerns rather than definitive fraud. Four methodological and reporting inconsistencies were identified in this 2023 neuroimaging paper on Wilson's disease (DOI: 10.1007/s11682-022-00753-3). First, the mean ages of WD patients (22.39) and healthy controls (22.78) in the text are inconsistently summarized as 23 in Table 1, suggesting careless data handling. Second, Table 1 reports EBPM values as '2–8 (4.74–1.52)' instead of the conventional '4.74 ± 1.52', an apparent transcription error. Third, the authors state they used MATLAB R2012a for AFQ analysis in a 2023 publication, which is unusually outdated and may indicate methodological copy-pasting from an earlier study. Fourth, a one-way ANOVA is used for two-group comparisons, where independent-sample t-tests would be standard, suggesting template reuse. No image manipulation or fabricated datasets were detected. These issues point to lax proofreading and possible methodological copying rather than deliberate fraud. Confidence is moderate; confirmation requires author response and raw data review.

Verdict

Questionable — The paper contains multiple reporting and methodological inconsistencies that fall short of outright fraud but indicate inadequate manuscript preparation and possible template reuse. Independent verification from authors and editors is recommended.

Key findings

Age inconsistency between text and Table 1: WD mean age given as 22.39 (SD = 6.35) and HC as 22.78 (SD = 7.34) in the text, but Table 1 lists both groups as '23'.
Statistical symbol transcription error: EBPM row in Table 1 reads '2–8 (4.74–1.52)', which appears to be a misuse of a minus sign (–) for the plus–minus (±) operator; the intended value is '4.74 ± 1.52'.
Outdated software declaration: Methods state MATLAB R2012a was used for AFQ in a paper accepted in late 2022 / published in 2023, raising questions about either neglected laboratory infrastructure or direct copying of methods text from older publications.
Inappropriate statistical test for design: A one-way ANOVA is applied to a two-group (WD vs. HC) comparison, where independent-sample t-tests are standard practice. While mathematically equivalent (F = t²), this suggests methodological template copy-pasting.

Evidence highlights

Finding 1 (verified): Numerical discrepancy between in-text means (22.39 / 22.78) and the integer means shown in Table 1 (23 for both groups). The decimal values round to 22 and 23 respectively, not to a uniform '23' for both.
Finding 2 (verified): Table 1, EBPM row literal text: '2–8 (4.74–1.52)'. Standard convention is range (mean ± SD), i.e. '4.74 ± 1.52'.
Finding 3 (verified): Materials and methods explicitly cites MATLAB R2012a, released in 2012, used over a decade later for 3.0T MRI AFQ tractography.
Finding 4 (verified): Methods describe a one-way ANOVA applied to pointwise diffusion indicators between WD and HC groups.

Notes

DOI preserved as reported: 10.1007/s11682-022-00753-3.
No image manipulation, duplicated figures, fabricated data, or plagiarism of scientific content were detected within the scope of this report.
The findings suggest sloppiness or methodological copying rather than intentional data fabrication, but the cumulative pattern warrants editorial follow-up.
Confidence is moderate; raw data and reproducible code/software logs from the authors would help resolve the software-version and statistical-test questions.
This report was AI-assisted and is intended for academic discussion only; institutional investigation is required for a final determination.