Medical Image Harmonization Using Deep Learning Based Canonical Mapping: Toward Robust and Generalizable Learning in Imaging
This addresses the challenge of clinical adoption for medical imaging AI by improving robustness across varied acquisition conditions, though it is incremental in applying existing generative models to a known bottleneck.
The paper tackles the problem of limited generalization in medical imaging models across different devices and protocols by proposing a deep learning-based harmonization method that maps diverse data to a common reference domain. Results show substantial improvement in out-of-sample performance for MRI-based brain age prediction and schizophrenia classification, using data from 9 sites and 9701 subjects, even with training restricted to a single site.
Conventional and deep learning-based methods have shown great potential in the medical imaging domain, as means for deriving diagnostic, prognostic, and predictive biomarkers, and by contributing to precision medicine. However, these methods have yet to see widespread clinical adoption, in part due to limited generalization performance across various imaging devices, acquisition protocols, and patient populations. In this work, we propose a new paradigm in which data from a diverse range of acquisition conditions are "harmonized" to a common reference domain, where accurate model learning and prediction can take place. By learning an unsupervised image to image canonical mapping from diverse datasets to a reference domain using generative deep learning models, we aim to reduce confounding data variation while preserving semantic information, thereby rendering the learning task easier in the reference domain. We test this approach on two example problems, namely MRI-based brain age prediction and classification of schizophrenia, leveraging pooled cohorts of neuroimaging MRI data spanning 9 sites and 9701 subjects. Our results indicate a substantial improvement in these tasks in out-of-sample data, even when training is restricted to a single site.