Automated Deep Learning-Based Demyelination Load Segmentation in Metachromatic Leukodystrophy

Abstract Purpose Metachromatic leukodystrophy (MLD) is a rare lysosomal storage disorder characterized by progressive white matter demyelination. Quantification of demyelinated white matter on MRI—typically expressed as the demyelination load —serves as a key imaging biomarker of disease burden, enabling objective monitoring beyond visual rating scales. However, current semi-automated pipelines are limited by manual interaction, pediatric brain variability, and differences in MRI acquisition. This study aimed to develop and validate a self-configuring convolutional neural network (CNN) for automated segmentation of demyelinated white matter in MLD and to compare its performance with a conventional semi-automated method across heterogeneous MRI datasets. Methods An nnU-Net was trained on 189 3D T1- and axial T2-weighted scans from 35 MLD patients using visually controlled conventional masks as ground truth. Independent testing was performed on 130 scans (73 high-resolution 3D, 57 lower-resolution 2D T1-weighted) from 49 patients. Performance was assessed by Dice coefficient, Bland-Altman bias, correlation with Gross Motor Function Classification (GMFC-MLD), MLD MRI severity score, longitudinal consistency, and qualitative review of outliers. Results CNN-based segmentation showed strong spatial agreement with the reference method, with a median Dice coefficient of 0.82 for 3D T1-weighted scans and 0.75 for 2D scans. Volumetric bias was minimal on Bland-Altman analysis. CNN-derived demyelination load correlated significantly with motor impairment (r S = 0.38 for 3D and r = 0.56 for 2D; both p < 0.001) and showed a stronger association with the MLD MRI severity score than conventional segmentation (3D: r S = 0.48 vs. 0.28; 2D: r S = 0.83 vs. 0.29). Correlations with clinical status were slightly lower (CNN: r S = 0.38, p < 0.001; conventional: (r S = 0.26, p < 0.025)) Longitudinal analyses demonstrated stable, monotonic changes over time, and qualitative review revealed fewer boundary misclassifications. Conclusion The nnU-Net enables fast, reproducible, and clinically meaningful segmentation of demyelinated white matter in MLD. It generalizes across MRI protocols, correlates with motor function, and offers a scalable tool for standardized biomarker extraction in clinical trials and other leukodystrophies.