Diagnose Harmonisation (DHARM): An open-source tool for the harmonisation and analysis of batch effects in MRI-derived measures.

DHARM, an open-source Python toolbox, provides standardized multi-criteria diagnostics showing that MRI harmonisation substantially reduces additive batch effects across UK Biobank, Cam-CAN, and OxBAC.

• Why it matters: Harmonisation validation often uses method- and dataset-specific criteria, making it hard to decide whether technical variation is sufficiently removed without losing biological signal.
• What they did: DHARM offers GUI/command-line workflows that quantify batch-associated differences in location, scale, empirical distributions, covariance, and multivariate structure, plus checks for biological covariate effects and repeated-measures consistency.
• The result: In cross-sectional multisite data, additive batch effects drop markedly, while residual variance and distribution differences depend on the harmonisation method; in repeated-measures travelling-heads, batch- and within-subject variability are reduced while subject-level and biological relationships are broadly preserved.

AI summary of the abstract · may contain errors · not medical advice · read the paper.