Accompanying pattern detection for tumor pathological section using statistical method

Computer assisted diagnosis has been widely used in practice with the fast development of medical imaging. As a key technique, medical image processing plays an essential role in computer assisted diagnosis. However, most pattern discovering methods focus on human structure images as CT, MRT while very little work has been done in micrograph. In this paper, we proposed a set of simple statistical operations to detect accompanying pattern in micrograph. The samples of different tissues are compared from both spatial distribution and growing direction to check if they fit in the same pattern. The main purpose of this research is to help cancer expert analyze the cause of cancer. Taken breast cancer as experimental case, our results illustrated the accompanying pattern between blood vessels and dead cells.