Web-Integrated AI System for Bone Cancer Prediction

Bone cancer detection from X-ray images is challenging due to subtle tumor characteristics, radiologist fatigue, and limited access to specialized expertise, particularly in resource-constrained settings. Early and accurate diagnosis is critical for improving patient outcomes. However, conventional methods rely heavily on manual interpretation, leading to delays and variability. This study proposes a web-integrated AI system for automated bone cancer prediction using deep learning. The system employs the EfficientNet-B0 architecture with transfer learning to classify bone X-ray images as normal or cancerous, supported by preprocessing, data augmentation, and class-weighting techniques to address data scarcity and class imbalance. The trained model is deployed as a user-friendly web application enabling X-ray upload, report generation, scan history management, and interaction through an integrated AI chatbot. The proposed system aims to provide an efficient, interpretable, and clinically supportive diagnostic tool for early bone cancer detection.