VisiHealth AI | End-to-End Medical VQA Demo | Flask Backend + Next.js Frontend + Explainable AI
Check out my Final Year Project – VisiHealth AI, a cutting-edge AI system that combines Computer Vision, NLP, and Explainable AI to answer clinical-style questions from medical images! 🩺🤖
💡 What VisiHealth AI Can Do:
Upload any medical image 🖼️
Ask natural language medical questions ❓
Get AI-generated answers with confidence scores ✅
See top predictions for better insight 🔍
Visualize attention maps and ROI highlights for explainability 🧠✨
Save and review analysis history 📂
💻 Tech Stack & AI Models:
Frontend: Next.js, React, TypeScript, Tailwind CSS, Framer Motion 🌐
Backend: Flask REST API ⚡
AI/ML: ResNet50 (medical images) + BioLinkBERT (NLP) + Multimodal Fusion + Knowledge Graph rationale 🧩
Dataset: SLAKE (Medical VQA) 📚
🎯 Project Highlights:
221 answer classes, with ≈61% validation accuracy 📊
Knowledge graph with 4,444 triplets for explainable reasoning 🧩
Interactive demo with fully explainable AI outputs 🔬
⚠️ Note: This is a research & educational project – not a clinical diagnostic tool.
💻 Explore More:
Portfolio: https://junaid-portfolio-ruby.vercel....
GitHub: https://github.com/juni2003
👍 Like, share, and subscribe for more AI demos!
💬 Feedback and questions welcome in the comments!
#MedicalAI #MedicalVQA #ExplainableAI #DeepLearning #ComputerVision #NLP #ResNet50 #BioLinkBERT #HealthcareAI #StudentProject #PortfolioProject #FinalYearProject #FlaskAPI #NextJS #ReactJS #TailwindCSS