Early Detection of Lung Cancer Using a Convolutional Neural Network Integrating Multidimensional Clinical Information

Lung cancer remains one of the leading causes of cancer-related deaths worldwide, primarily due to late-stage detection. Early diagnosis significantly improves survival rates, yet current screening methods face limitations in cost, accessibility, and predictive accuracy. This study leverages the comprehensive data from Taiwan's National Health Insurance Research Database (NHIRD) to develop a predictive model for early lung cancer detection. We propose a novel approach that integrates multidimensional clinical information, including diagnostic records, medication histories, and laboratory test results from the past three years, into a Multi-channel Convolutional Neural Network (MC-CNN). This architecture is designed to efficiently extract and analyze temporal and contextual patterns across diverse data modalities. Our model outperforms several comparative methods with a F1-score of 64%. These results underscore the potential of using advanced AI-driven approaches to facilitate early detection of lung cancer, enabling timely intervention and improved patient outcomes. This study not only demonstrates the efficacy of integrating multidimensional clinical data but also highlights the practical application of AI in addressing pressing challenges in healthcare.