Please join us for a CCN Seminar with Xiaoling Hu, PhD Candidate at Stony Brook University
Title: Topology-Informed Image Analysis
Abstract: In the segmentation of fine-scale structures from natural and biomedical images, per-pixel accuracy is not the only metric of concern. Topological correctness, such as vessel connectivity and membrane closure, is crucial for downstream analysis tasks. To address this issue, we propose new approaches to train deep image segmentation networks for better topological accuracy. Firstly, we design a continuous-valued loss function that enforces a segmentation to have the same topology as the ground truth, i.e., having the same Betti number. The proposed topology-preserving loss function is differentiable and we incorporate it into end-to-end training of a deep neural network. Moreover, leveraging the power of discrete Morse theory (DMT), we identify global structures, including 1D skeletons and 2D patches, which are important for topological accuracy. Trained with a novel loss based on these global structures, the network performance is significantly improved especially near topologically challenging locations (such as weak spots of connections and membranes). To overcome the issues caused by pixel-wise predictions, we further propose the first deep learning based method to learn topological/structural representations. And then we learn a probabilistic model that can perform inference tasks in such a topological/structural representation space. On diverse datasets, our method achieves superior performance on both the DICE score and topological metrics.