Method and system for segmenting overlapping cytoplasms in medical image

Abstract

A method for segmenting overlapping cytoplasms in a medical image, comprising: establishing a cytoplasm shape hypothesis set (201); selecting a shape hypothesis for each cytoplasm from within the established cytoplasm shape hypothesis set to perform constrained multi-shape evolution (202), thus segmenting overlapping cytoplasm in a medical image, wherein the constrained multi-shape evolution (202)comprises: segmenting a clump region composed of a plurality of overlapping cytoplasms so as to provide clump evidence (301); aligning the shape used to evaluate the quality of the selected shape hypothesis (302); determining the shape evolution of a better shape hypothesis for each cytoplasm (303). Also relating to a system for segmenting overlapping cytoplasms in a medical image. The present system and method may more accurately and effectively segment overlapping cytoplasms on the basis of prior shapes. 一种分割医学图像中的重叠细胞质的方法，其包括：建立细胞质形状假设集(201)；以及从建立的细胞质形状假设集中为每个细胞质选择形状假设，来进行受约束的多形状演进(202)，从而分割医学图像中的重叠细胞质，其中，受约束的多形状演进(202)包括：对由多个重叠细胞质构成的团块区域进行分割以提供团块证据(301)；用于评估选择的形状假设的质量的形状对齐(302)；以及为每个细胞质确定更好的形状假设的形状演进(303)。还涉及一种分割医学图像中的重叠细胞质的系统。系统和方法能够更精确、更有效地基于先验形状进行重叠细胞质的分割。