Artificial Intelligence for Secure Remote Healthcare Monitoring
Developed a prototype integrating an ensemble ML model with MedGemma 4b-it for disease prediction and clinical triage recommendations. Achieved 95.2% accuracy and F1 score across 5 disease classes: asthma, diabetes mellitus, healthy, heart disease, and hypertension. Built an interactive Streamlit Cloud app enabling patient vitals input, risk analysis, and AI-guided triage feedback.