CDNet: Centripetal Direction Network for Nuclear Instance Segmentation

Nuclear instance segmentation is a challenging task due to a large number of touching and overlapping nuclei in pathological images. Existing methods cannot effectively recognize the accurate boundary owing to neglecting the relationship between pixels (e.g., direction information). In this paper, we propose a novel Centripetal Direction Net-work (CDNet) for nuclear instance segmentation. Specifically, we define centripetal direction feature as a class of adjacent directions pointing to the nuclear center to rep-resent the spatial relationship between pixels within the nucleus. These direction features are then used to construct a direction difference map to represent the similarity within instances and the differences between instances. Finally, we propose a direction-guided refinement module, which acts as a plug-and-play module to effectively integrate auxiliary tasks and aggregate the features of different branches. Experiments on MoNuSeg and CPM17 datasets show that CDNet is significantly better than the other methods and achieves the state-of-the-art performance. The code is available at https://github.com/honglianghe/CDNet.

1. Introduction

Nuclear instance segmentation techniques enable accurate quantitative characterizations of nuclear size and shape (e.g., circularity and aspect ratio), which are key components of the study of cancerous tissues [9]. However, a whole-slide image contains tens of thousands of nuclei of various types (as shown in Fig. 1(a)∼(b)), and nuclei display a great deal of inter- and intra-instance variability because of their appearances, surroundings by organs, disease types, and even digital scanner brands. In particular, tumour nuclei tend to be present in clusters and lead to clustered overlapping instances, which also provides challenge for accurate segmentation of nuclear instances.