Advancing colorectal cancer detection with inception-residual convolutional neural networks

Gastrointestinal (GI) tract diseases pose significant challenges to medical professionals due to their complex nature and diverse manifestations. This study presents a novel approach to tackle these challenges by integrating advanced deep learning techniques. The proposed method employs an Inception-Residual based Deep Convolutional Neural Network (IRDCNN) for the automated classification of colorectal cancer. Initially, the input image undergoes contrast enhancement using Contrast Limited Adaptive Histogram Equalization (CLAHE) to enhance the visibility of infected regions. Through extensive training on annotated datasets, proposed model learns to identify the colorectal cancer with high accuracy and reliability. By automating the classification process, the proposed approach streamlines the diagnostic workflow, enabling faster and more efficient detection of colorectal cancer. Overall, the proposed IRDCNN-based framework represents a promising avenue for enhancing the diagnosis of colorectal cancer with achieving accuracy of 96.8%, ultimately contributing to improved healthcare delivery and patient care.