Feather-Light Fourier Domain Adaptation in Magnetic Resonance Imaging
This addresses domain shift issues in medical imaging for clinics, though it is incremental as it builds on existing Fourier-based adaptation ideas.
The paper tackles the problem of domain shift in MRI data from different hardware by proposing a lightweight, training-free test-time adaptation method that replaces low-frequency Fourier components of target images with those from source data, achieving state-of-the-art performance comparable to complex deep models.
Generalizability of deep learning models may be severely affected by the difference in the distributions of the train (source domain) and the test (target domain) sets, e.g., when the sets are produced by different hardware. As a consequence of this domain shift, a certain model might perform well on data from one clinic, and then fail when deployed in another. We propose a very light and transparent approach to perform test-time domain adaptation. The idea is to substitute the target low-frequency Fourier space components that are deemed to reflect the style of an image. To maximize the performance, we implement the "optimal style donor" selection technique, and use a number of source data points for altering a single target scan appearance (Multi-Source Transferring). We study the effect of severity of domain shift on the performance of the method, and show that our training-free approach reaches the state-of-the-art level of complicated deep domain adaptation models. The code for our experiments is released.