SIGMO: Sigmoid Colon Cancer Prediction Using Machine Learning and Deep Learning

Many sigmoid colon cancer prediction systems rely on traditional methods, which are prone to human error and ineffective in spotting minute variations. However, systems based on Random Forest (RF) and Support Vector Machines (SVM) are known to have drawbacks, including poor generalization and overfitting. Predictive Model machine learning (ML) and deep learning (DL) models such as Convolutional Neural Networks (CNNs), autoencoders, and XGBoost are used in the proposed SIGMO to address these issues. It also integrates rich and unstructured medical images and robust and structured clinical data to provide accurate and dependable predictions. Therefore, SIGMO exhibits cutting-edge efficiency by utilizing feature extraction, dimensionality reduction, and ensemble approaches, which clarifies the reasons for the underperformance of existing systems. These results demonstrate the significance of integrating multi-modal data for diagnosis and the real-time deployment of such systems in addition to enhanced diagnostic accuracy. It also shows a significant increase (92.8%) in accuracy, precision (89.6%), recall (94.2%), and AUC-ROC (95%) over traditional systems..