One-Shot Segmentation of Novel White Matter Tracts via Extensive Data Augmentation

Abstract

AbstractDeep learning based methods have achieved state-of-the-art performance for automated white matter (WM) tract segmentation. In these methods, the segmentation model needs to be trained with a large number of manually annotated scans, which can be accumulated throughout time. When novel WM tracts—i.e., tracts not included in the existing annotated WM tracts—are to be segmented, additional annotations of these novel WM tracts need to be collected. Since tract annotation is time-consuming and costly, it is desirable to make only a few annotations of novel WM tracts for training the segmentation model, and previous work has addressed this problem by transferring the knowledge learned for segmenting existing WM tracts to the segmentation of novel WM tracts. However, accurate segmentation of novel WM tracts can still be challenging in the one-shot setting, where only one scan is annotated for the novel WM tracts. In this work, we explore the problem of one-shot segmentation of novel WM tracts. Since in the one-shot setting the annotated training data is extremely scarce, based on the existing knowledge transfer framework, we propose to further perform extensive data augmentation for the single annotated scan, where synthetic annotated training data is produced. We have designed several different strategies that mask out regions in the single annotated scan for data augmentation. To avoid learning from potentially conflicting information in the synthetic training data produced by different data augmentation strategies, we choose to perform each strategy separately for network training and obtain multiple segmentation models. Then, the segmentation results given by these models are ensembled for the final segmentation of novel WM tracts. Our method was evaluated on public and in-house datasets. The experimental results show that our method improves the accuracy of one-shot segmentation of novel WM tracts.KeywordsWhite matter tract segmentationOne-shot learningData augmentation