Optimal Deep Transfer Learning Based Colorectal Cancer Detection and Classification Model

Abstract

Colorectal carcinoma (CRC) is one such dispersed cancer globally and also prominent one in causing cancer-based death. Conventionally, pathologists execute CRC diagnosis through visible scrutinizing under the microscope the resected tissue samples, stained and fixed through Haematoxylin and Eosin (H&E). The advancement of graphical processing systems has resulted in high potentiality for deep learning (DL) techniques in interpretating visual anatomy from high resolution medical images. This study develops a slime mould algorithm with deep transfer learning enabled colorectal cancer detection and classification (SMADTL-CCDC) algorithm. The presented SMADTL-CCDC technique intends to appropriately recognize the occurrence of colorectal cancer. To accomplish this, the SMADTL-CCDC model initially undergoes pre-processing to improve the input image quality. In addition, a dense-EfficientNet technique was employed to extract feature vectors from the pre-processed images. Moreover, SMA with Discrete Hopfield neural network (DHNN) method was applied for the recognition and classification of colorectal cancer. The utilization of SMA assists in appropriately selecting the parameters involved in the DHNN approach. A wide range of experiments was implemented on benchmark datasets to assess the classification performance. A comprehensive comparative study highlighted the better performance of the SMADTL-CDC model over the recent approaches.