A novel geodesic distance based clustering approach to delineating boundaries of touching cells
ISNN'13 Proceedings of the 10th international conference on Advances in Neural Networks - Volume Part II
| Hi-index | 0.00 |
Automatic cell segmentation and dead cell detection in microscopic images play a very important role in the study of the behaviour of lymphocytes. In this paper, a distance and watershed transforms based cell segmentation algorithm has been proposed to segment cells by using CFSE image, and a dead cell detection algorithm is also proposed to detect cell dead event. Experimental results have shown that the proposed algorithms are pretty robust to variable contrast microscopy image data, and variable cell densities, and the average cell detection rate has reached 93% with the average miss detection rate about 7%, and extremely low average false detection rate of 0.7%, and the dead cell rate is about 11%.