A Novel Fusion Approach for Early Lung Cancer Detection Using Computer Aided Diagnosis Techniques

Computer Aided Diagnosis (CAD) plays an effective and important role in radiology. It provides second opinion to the radiologists during patient image assessment. In this work, early detection of lung cancer is chosen for the study. Computed tomography is one of the normally preferred modality to record the interior body parts, particularly lungs. Recent advances in radiology also supports to record the two dimensional (2D) and three dimensional (3D) images of lungs which is associated with the abnormality, such as lesion/tumor. The main clinical challenge is to develop a suitable CAD system and Content Supported Medical Image Retrieval (CSMIR) system to extract and analyze the lesion/tumor from 2D and 3D radiology images. Hence, it is essential to develop an automated system with the following capability: detection, categorization and quantification of the lung lesion/tumor. In the proposed work, lung abnormality is segmented using the wavelet approach and the segmented Region of Interest (ROI) is then classified using a novel classifier unit. The experimental result confirms that, proposed approach offers enhanced accuracy and specificity compared with the other methods considered in this study.