Clinical Decision Support System for Lung Cancer Diagnosis Using Convolutional Neural Networks

Abstract

According to WHO, lung cancer is the leading cause of cancer-related deaths worldwide. A study has found that early detection of lung cancer using a patient's CT scans has been effective in reducing the deaths caused by lung cancer. Through the use of convolutional neural networks with CT scans as input, a clinical decision support system for lung cancer diagnosis is developed to aid doctors that are non-radiologists in classifying if a patient is positive or negative for lung cancer. The current model used by the system needs to be further enhanced before being deployed for use by doctors. If the model is improved, it could be helpful in providing second opinion on detecting lung cancer in patients.