Diagnosis of Lung Cancer Based on CT Scans Using CNN
Hamdalla F. Al-Yasriy, Muayed S. AL-Husieny, Furat Y. Mohsen, Enam A. Khalil, Zainab S. Hassan
Abstract
Abstract Lung cancer is one of the most lethal cancer types; thousands of peoples are infected with this type of cancer, and if they do not discover it in the early stages of the disease, then the chance of surviving of the patient will be very poor. For the suggested reasons above and to help in overcoming this terrible, early diagnosis with the assistance of artificial intelligence procedures most needed. Through this research, a Computer-aided system introduced for detecting lung cancer in a dataset collected from the Iraqi hospitals by using a convolutional neural network technique with AlexNet architecture for helping with the diagnosis of the patient’s cases: normal, benign, or malignant. The proposed model gives high accuracy ups to 93.548%. The other performance metrics comes with high values such as 95.714% for sensitivity and 95% for Specificity.