Segmentation of Cellular Structures with Encoder-Decoder Based Deep Learning Algorithm in Histopathological Images

Abstract

In this proposed study, SegNet is used for segmentation of cellular structures in high-resolution histopathological images. SegNet is a deep convolutional encoder-decoder architecture used for segmenting objects on the road and indoors. In this proposed study, the segmentation performance of SegNet algorithm for cellular structures in high resolution histopathological images is tried to be obtained. We also compared the performance of the SegNet algorithm with the state-of-the-art segmentation algorithms in the literature. According to the obtained results, SegNet has been observed to be quite successful compared to other methods commonly used in the segmentation of cellular structures.