A Novel Approach for Evaluating the Performance of a Computer-Assisted System for the Diagnosis of Lung Cancer Using a Combination of Image Processing and Machine Learning Techniques
Authors: J. Zhang, L. Wang, Y. Chen, H. Liu, X. Zhao
Lung cancer remains a leading cause of cancer-related mortality worldwide, and early detection is crucial for improving patient outcomes. Computer-aided diagnosis (CAD) systems have shown promise in assisting radiologists with the interpretation of medical images. In this study, we propose a novel CAD system that integrates advanced image processing techniques with a deep learning-based classifier to automatically detect and classify pulmonary nodules in computed tomography (CT) scans. The system employs a multi-scale feature extraction module and a hybrid attention mechanism to enhance the discriminative power of the model. We evaluated the system on a large dataset of CT images from multiple institutions, achieving a sensitivity of 94.2% and a specificity of 91.8%, with an area under the receiver operating characteristic curve (AUC) of 0.97. The proposed method outperformed several state-of-the-art approaches in terms of accuracy and robustness. Our findings suggest that the integration of sophisticated image processing and machine learning techniques can significantly improve the diagnostic performance of CAD systems for lung cancer, potentially aiding in earlier detection and reducing unnecessary biopsies.
