Thesis (M.Sc.) - Cairo University - Faculty of Engineering - Department of Systems and Biomedical Engineering

Lung cancer remains the leading cause of cancer-related deaths in the world. Early diagnosis can improve the effectiveness of treatment and increase the patient{u201F}s chance of survival. Detecting pulmonary nodules in chest computed tomography (CT) scans helps to diagnose lung cancer in an early stage, assess the risk of malignancy. A computer-aided detection (CADe) system is developed by proposing algorithms for classification of lung tissues in chest CT scans to distinguish between nodules (abnormal tissues) and non-nodule (normal tissues) using different combination of 2-D features. The components of the CADe system include feature extraction, feature normalization, feature selection and binary classification (Nodule or Non-Nodule). The system provided excellent results with high sensitivity and specificity on data obtained from two different international reference data sets

Issued also as CD