Optimized CNN Architecture Using Bio-Inspired and Computing for Lung Cancer Detection in Infrared Images

Lung cancer is arguably the most rampant cancer that causes deaths across the globe, and early diagnosis is a key factor in elevating survival. This work introduces an improved design of CNN and other five bio-inspired computational algorithms: AGO, SDFS, BSGB, DRS, and HEL. These midlevels were designed and optimized for the true detection of lung cancer in infrared images with a high degree of accuracy of 97% for differentiating between malignant and benign diseases. Our system synthesizes pattern recognition with decision support using heterogeneous data including CT scans, chest X-rays, and patient clinical history, enabling quick and accurate diagnoses. The design of such a system may well transform the global health field through improved early detection rates, decreased diagnostic costs, and prompt treatments. The discovery leads not only to arming clinicians with decision-making tools but also to creating the potential for AI-based cancer treatment and prevention to save millions of people's lives around the globe.