Design of Deep Learning-based Approach to Predict Lung Cancer on CT Scan Images

This study presents the design of a deep learning-based approach for predicting lung cancer from computerized tomography (CT) scan images. Lung cancer is one of the major global health concern, and early detection is critical for improving patient outcomes. This study aims to develop an accurate and efficient model capable of detecting lung cancer from CT scans by using deep learning algorithms.The proposed method involves preprocessing of CT images to improve their quality, followed by feature extraction using deep convolutional neural networks. These characteristics are then fed into a classification model, which predicts whether or not lung cancer exists. The model’s performance will be analysed using a large dataset of CT scans, which will include both cancerous and non-cancerous cases with various stages of malignancy. The proposed model is also deployed on Raspberry Pi to detect the cancer related abnormalities in the CT scan images. This study managed to achieve an average accuracy of 72.41% in classification of lung cancer from CT images. The results of this study have the potential to provide clinicians with a dependable tool for detecting early lung cancer, allowing for rapid treatment and improved patient care.