PGL: Prior-Guided Local Self-supervised Learning for 3D Medical Image Segmentation
This work provides an incremental improvement for medical image segmentation, specifically for 3D data where dense annotations are scarce, by focusing on local feature consistency.
This paper addresses the challenge of 3D medical image segmentation with limited annotations by proposing a Prior-Guided Local (PGL) self-supervised learning model. PGL learns region-wise local consistency in the latent feature space, leveraging spatial transformations to align feature maps of local regions across augmented views. This pre-training method significantly improves segmentation performance on four CT datasets covering 11 organs and two tumors compared to random and global consistency-based initializations.
It has been widely recognized that the success of deep learning in image segmentation relies overwhelmingly on a myriad amount of densely annotated training data, which, however, are difficult to obtain due to the tremendous labor and expertise required, particularly for annotating 3D medical images. Although self-supervised learning (SSL) has shown great potential to address this issue, most SSL approaches focus only on image-level global consistency, but ignore the local consistency which plays a pivotal role in capturing structural information for dense prediction tasks such as segmentation. In this paper, we propose a PriorGuided Local (PGL) self-supervised model that learns the region-wise local consistency in the latent feature space. Specifically, we use the spatial transformations, which produce different augmented views of the same image, as a prior to deduce the location relation between two views, which is then used to align the feature maps of the same local region but being extracted on two views. Next, we construct a local consistency loss to minimize the voxel-wise discrepancy between the aligned feature maps. Thus, our PGL model learns the distinctive representations of local regions, and hence is able to retain structural information. This ability is conducive to downstream segmentation tasks. We conducted an extensive evaluation on four public computerized tomography (CT) datasets that cover 11 kinds of major human organs and two tumors. The results indicate that using pre-trained PGL model to initialize a downstream network leads to a substantial performance improvement over both random initialization and the initialization with global consistency-based models. Code and pre-trained weights will be made available at: https://git.io/PGL.