25. MACHINE VISION AND AI INTEGRATION FOR BLOODLESS HEALTH PROFILING

Traditional blood tests are invasive, time-consuming, and require laboratory facilities, making them unsuitable for continuous monitoring and large-scale screening. To address these limitations, this study proposes an AI-powered, camera-based bloodless health assessment system. The method uses remote photoplethysmography (rPPG) to capture vascular and chromatic changes from the skin, with PCA and ICA applied for noise reduction. Extracted features are analyzed using a hybrid CNN-LSTM model, integrating spatial and temporal data for accurate biomarker estimation. Transfer learning with MobileNetV2 and ResNet further improves efficiency and accuracy. The system estimates parameters such as glucose, hemoglobin, cholesterol, and oxygen saturation, generating a health report without invasive sampling. A conversational Chatbot Assistant is integrated to guide data capture, explain results, answer user queries, and recommend next steps with safety guardrails. Initial results show promising accuracy, highlighting its potential as a cost-effective, scalable tool for preventive healthcare and population screening.