Automatic lung nodule detection system using image processing techniques in computed tomography

Abstract

Listen

Translate article

Diagnosing and treating lung cancer at an early stage can improve the survival rate of patients. This study attempted to develop a computer-aided detection (CAD) system. In order to include all nodule types in the detection, this study proposes an image processing method for detecting ground glass opacity (GGO), part solid, and solid nodules in chest computed tomography. The process comprises image preprocessing, lung segmentation, nodule enhancement, candidate detection, and reduction of false positives. For lung segmentation, the edge searching method replaces the computing-intensive iterative hole-filling method. In order to extract nodules with extensively distributed gray levels, image accumulation is used in the nodule enhancement to rapidly enhance the gray level of individual nodules. In order to reduce false positives, the support vector machine (SVM) is applied twice. On the first run, the candidate nodules are obtained by using 4 two-dimensional features, and the classification result is obtained by using 11 three-dimensional features on the second run. This study used 667 lung nodules for experiment and evaluation. The proposed system can detect GGO, part solid, and solid nodules and takes only 0.1 s to process a single image. The total sensitivity of the system is more than 92.05%. The system excels at detecting small nodules in the range of 5 mm–9 mm with a sensitivity of 93.73% and GGO with a sensitivity of 93.02%. The results showed that the proposed rapid detection system has high sensitivity and low false positives, contributing to helping the clinicians’ diagnosis.