$\texttt{DisMorph}$: learning to disentangle technical distortions from true biological change

2026-08-08

AI Summary

In longitudinal MRI, image registration struggles to distinguish geometric distortions caused by technical factors—such as gradient nonlinearity—from genuine anatomical changes, leading to biased morphometric measurements. To address this, this work proposes DisMorph, a novel framework that explicitly decouples technical distortions from biological structural changes during registration. DisMorph employs a generative model trained on synthetic data to predict two separate dense deformation fields corresponding to each component, thereby achieving disentanglement. Domain randomization is further integrated to enhance generalization across imaging protocols. Experiments demonstrate that the method more accurately detects anatomical changes in simulated data, correctly attributes alterations in real image pairs containing only distortion, and effectively identifies both disease-related structural changes and residual distortions in longitudinal Alzheimer’s disease datasets.

Abstract

Longitudinal MRI enables sensitive measurement of structural brain change for studying aging and neurodegenerative disease. Deformable image registration is a key tool for estimating such change by computing a dense deformation that captures geometric differences between longitudinal scans. However, MRI scanners introduce geometric distortions that vary across acquisition systems and protocols, such as gradient non-linearity (GNL) distortion. Existing registration methods estimate a single field that conflates biological and technical effects, potentially biasing downstream morphometric measurements if distortions remain (partially) uncorrected. We propose $\texttt{DisMorph}$, a registration framework trained entirely on synthetic data that explicitly decomposes longitudinal deformation into technical and anatomical transforms. It predicts two dense deformations, each encoding one effect. During training, a novel generative model synthesizes both effects separately to provide disentanglement supervision, while domain randomization promotes generalization across imaging protocols. We evaluate our method in three complementary settings. On simulated data with known ground truth, our method detects anatomical change more accurately and consistently than conventional registration. On real image pairs that differ only by GNL distortion, our method assigns most geometric change to the distortion field, demonstrating specificity in the absence of anatomical change. On longitudinal Alzheimer's disease (AD) pairs, our method detects anatomical change in AD-related brain structures while identifying residual distortion left after standard correction. By disentangling MRI-induced distortion from biological change in the longitudinal deformation, our method paves the way for more accurate longitudinal morphometry in clinical settings where maintaining acquisition consistency is challenging.

Problem

• longitudinal MRI
• geometric distortion
• image registration
• morphometry
• disentanglement

Innovation

• disentanglement
• longitudinal MRI
• geometric distortion
• deformable registration
• synthetic data