A Hybrid Deep Learning Architecture Integrating YOLOv8 and ResNet for Lung Disease Diagnosis

Abstract

This project introduces an AI-powered diagnostic platform designed to identify major respiratory conditions from chest X-ray images using a advanced hybrid Deep Learning architecture. By integrating YOLOv8 for precise lesion localization and ResNet50 for deep feature extraction, the system overcomes the limitations of traditional single-model approaches, offering a more detailed analysis of lung pathology. The model is trained on a comprehensive dataset encompassing four critical categories: Normal, COVID-19, Pneumonia, and Tuberculosis. To ensure clinical reliability, the system employs advanced preprocessing including normalization and augmentation to handle variations in X-ray quality. This dual-network engine is integrated into a responsive web application that provides healthcare providers with near-instantaneous diagnostic results and confidence scores. With a user-friendly interface designed for both specialists and general practitioners, the platform bridges the gap in medical expertise, particularly in resource-limited or remote regions. By combining automated detection with accessible web technology, this research provides a scalable solution to accelerate clinical decision-making and improve patient outcomes in respiratory healthcare.