Improved Moth Flame Optimization with Deep Convolutional Neural Network for Colorectal Cancer Classification using Biomedical Images
Parimala S, R. Kumutha, Faiza Iram, Sunena Rose M V, R. Nanmaran, A. Angelpreethi · 2024
Colorectal cancer (CRC) affects the large intestinal colon. CRC treated radiation, chemotherapy, and immunotherapy. If aberrant cells, which are predicted to become cancer cells, are detected early, CRC may be effectively managed. Histopathological analysis of tissue specimens is the most common way to diagnose CRC. Histopathological examination and appropriate therapy are the only ways to improve cancer survival. AI in colon and lung cancer histopathology analysis may help doctors identify instances faster, cheaper, and easier. Medical image processing may use ML devices for pattern recognition and resolution. This paper presents an Improved Moth Flame Optimisation with Deep Convolutional Neural Network (IMFO-DCNN) method for biomedical image-based colorectal cancer classification. The IMFO-DCNN approach uses the DCNN model to automatically extract biomedical picture properties. The IMFO method fine-tunes DCNN model hyperparameters to improve detection performance. The last step uses a Random Forest (RF) classification algorithm to effectively recognise and categorise biological pictures. Methodically estimating the IMFO-DCNN methodology on biological pictures improves CRC classification accuracy and resilience over traditional techniques. CRC detection and classification are efficient and reliable using IMFO-DCNN.