Mmt-Vin: An Intelligent Lung Cancer Detection Framework Utilizing Machine Learning

Lung cancer, one of the top causes of cancer-related deaths, relies on early detection for effective treatment. We suggest an automated system that leverages machine learning to analyze CT scans, assisting radiologists in accurately detecting suspicious lung nodules. The approach involves pre-processing CT images to highlight nodule features. Machine learning models, such as CNNs, are trained on annotated CT scans to recognize patterns of malignant and benign nodules. Different CNN architectures, including ResNet, DarkNet, and EfficientNet, are evaluated to determine the most efficient model for detection. We introduce a model called MMT-VIN, which is built upon the VGG-19 approach. The proposed lung cancer detection method achieves an accuracy of approximately 98.8%, outperforming the methods it was compared against.