A YOLO12s-based Blueprint and Experimental Study for AI-Assisted Pulmonary Nodule Detection

Enrique Flores Tenorio J.D.; Andre Goyas Ayllon L.; Reyes Silva P.D. · ICCIDS 2026 - 9th International Conference on Computational Intelligence in Data Science (2026) · DOI: 10.1109/iccids69108.2026.11407765

Resumen

Early detection of lung cancer remains a major challenge in the medical field, with a critical impact on patient survival rates. This paper proposes an innovative solution that integrates an artificial intelligence model into a web application, leveraging convolutional neural networks for the early detection of pulmonary nodules in computed tomography scans. This study aims to accelerate and optimize the detection process through an intelligent and accessible tool for medical imaging specialists.The solution employs deep learning techniques applied to YOLO architecture, which is extensively recognized for its object detection capabilities, selecting the YOLO12s version for initial implementation. In the experimental phase, the performance of the YOLO12s model achieved a precision of 0.960 and a recall of 0.910 on the LIDC-IDRI test set, demonstrating the capacity to minimize false positives and improve efficiency. This proposal provides an innovative and accessible alternative to support early detection of cancer, with potential applications in healthcare.

Instituciones: Universidad Peruana de Ciencias Aplicadas