Histogram Matching Augmentation for Domain Adaptation with Application to Multi-Centre, Multi-Vendor and Multi-Disease Cardiac Image Segmentation
This work provides an incremental solution for improving the robustness of CNN-based cardiac image segmentation across different MRI scanners and clinical centers for medical practitioners.
This paper addresses the performance degradation of Convolutional Neural Networks (CNNs) in cardiac image segmentation when training and testing data come from different domains. The authors propose a histogram matching (HM) data augmentation method to bridge this domain gap, achieving average Dice scores of 0.9051, 0.8405, and 0.8749 for different cardiac structures and ranking third in the MICCAI 2020 M&Ms challenge.
Convolutional Neural Networks (CNNs) have achieved high accuracy for cardiac structure segmentation if training cases and testing cases are from the same distribution. However, the performance would be degraded if the testing cases are from a distinct domain (e.g., new MRI scanners, clinical centers). In this paper, we propose a histogram matching (HM) data augmentation method to eliminate the domain gap. Specifically, our method generates new training cases by using HM to transfer the intensity distribution of testing cases to existing training cases. The proposed method is quite simple and can be used in a plug-and-play way in many segmentation tasks. The method is evaluated on MICCAI 2020 M\&Ms challenge, and achieves average Dice scores of 0.9051, 0.8405, and 0.8749, and Hausdorff Distances of 9.996, 12.49, and 12.68 for the left ventricular, myocardium, and right ventricular, respectively. Our results rank the third place in MICCAI 2020 M\&Ms challenge. The code and trained models are publicly available at \url{https://github.com/JunMa11/HM_DataAug}.