LUNG CANCER DETECTION AND PREDICTION USING DEEP LEARNING

Abstract

Lung cancer is a malignant tumor of the lungs that is characterized by uncontrolled cell proliferation. Early identification of lung cancer can improve lung cancer patient’s chances of survival and save many lives. Because cancer diagnosis is one of the most difficult tasks for radiologists due to the shape of cancer cells, a computer- aided diagnosis method can be beneficial. As a result, early detection and prediction of lung cancer should play a critical role in the diagnosis process, as well as increase patient survival rates. This project presents lung cancer detection based on CT images using efficient lung cancer classification and prediction using deep learning models for improved accuracy. The architecture is trained using preprocessed CT images. The patient input photos are then tested using deep learning models. The primary goal of this research is to determine whether a patient's lung containsa cancer tumor. By creating a handcrafted CNN model and using Transfer Learning-based VGG-16 and Inception V3 architectures to train the model, we present a convolution neural network-based classification technique. The performance of these models is compared in this study. To achieve the best outcomes, hyper parameter optimization was performed.