Accurate Nuclear Segmentation with Center Vector Encoding
This addresses the problem of accurate nuclear segmentation for pathology image analysis, which is incremental as it builds on existing methods with a new approach.
The paper tackles the challenge of nuclear segmentation in pathology images by introducing a novel bottom-up method using Center Mask and Center Vector concepts, which simplifies instance differentiation and outperforms state-of-the-art methods by a clear margin.
Nuclear segmentation is important and frequently demanded for pathology image analysis, yet is also challenging due to nuclear crowdedness and possible occlusion. In this paper, we present a novel bottom-up method for nuclear segmentation. The concepts of Center Mask and Center Vector are introduced to better depict the relationship between pixels and nuclear instances. The instance differentiation process are thus largely simplified and easier to understand. Experiments demonstrate the effectiveness of Center Vector Encoding, where our method outperforms state-of-the-arts by a clear margin.