Detection of Colorectal Cancer from Histopathological Images Tissue Classification Using Deep Learning Techniques

Abstract

Colon cancer is still a major public health issue, and improving patient outcomes depends critically on early and precise diagnosis. In this work, we introduce classifying colon cancer tissue by utilizing deep learning and two cutting-edge convolutional neural networks (CNNs), VGG16 and MobileNetV2. Adipose (ADI), Background (BACK), Debris (DEB), Lymphocytes (LYM), Mucus (MUC), Smooth Muscle (MUS), Normal Colon Mucosa (NORM), Cancer-Associated Stroma (STR), and Colorectal Adenocarcinoma Epithelium (TUM) are the tissue types on which our research focuses.We can accurately detect and categorize various tissue types automatically by using deep learning models, such as VGG16 and MobileNetV2. Our findings suggest that deep learning methods, specifically VGG16 and MobileNetV2, hold significant promise for enhancing tissue categorization accuracy in colon cancer diagnosis. In VGG16, we obtained an accuracy of 95%, while in MobileNet-v2, we obtained 97% accuracy. Additionally, another CNN model gave 87% accuracy in classifying colon cancer tissue. The results show that deep learning models can lead to better patient outcomes by recognizing tissue classes more rapidly and accurately, which in turn can lead to better colon cancer diagnosis.