The detection of abnormal lesions in the early stages of lung cancer is important to improve survival. Computer-aided detection (CAD) system can be useful for early detection of pulmonary nodules on computed tomography (CT) images for screening. Moreover, CAD system can be 'second opinion' when a radiologist detects the pulmonary nodules on multi-slice CT images. We developed a computer-aided detection system for pulmonary nodule detection on multi-detector row computed tomography (MDCT) images. We applied the nodule isolation technique using radial distribution function and additional algorithms. In this paper, we reported the ground-glass opacity (GGO) lesions detected by self-developed computer-aided pulmonary nodule detection system.