Harnessing Uncertainty in Domain Adaptation for MRI Prostate Lesion Segmentation
This work addresses the challenge of adapting medical imaging models to novel, more informative modalities, which is crucial for advancing cancer characterization in clinical settings, though it is incremental as it builds on existing domain adaptation techniques.
The paper tackled the problem of domain adaptation for MRI prostate lesion segmentation when translating from a less informative source domain (mp-MRI) to a richer target domain (VERDICT MRI), where multiple target samples can arise from a single source sample. The result was that explicitly modeling the uncertainty in this mapping and generating multiple outputs per input led to systematically better image representations, yielding substantial improvements over deterministic methods across various dataset sizes and baselines.
The need for training data can impede the adoption of novel imaging modalities for learning-based medical image analysis. Domain adaptation methods partially mitigate this problem by translating training data from a related source domain to a novel target domain, but typically assume that a one-to-one translation is possible. Our work addresses the challenge of adapting to a more informative target domain where multiple target samples can emerge from a single source sample. In particular we consider translating from mp-MRI to VERDICT, a richer MRI modality involving an optimized acquisition protocol for cancer characterization. We explicitly account for the inherent uncertainty of this mapping and exploit it to generate multiple outputs conditioned on a single input. Our results show that this allows us to extract systematically better image representations for the target domain, when used in tandem with both simple, CycleGAN-based baselines, as well as more powerful approaches that integrate discriminative segmentation losses and/or residual adapters. When compared to its deterministic counterparts, our approach yields substantial improvements across a broad range of dataset sizes, increasingly strong baselines, and evaluation measures.