AI-Assisted Diagnostic Support Tool for Common Skin Conditions
Dermatological conditions are among the most common health complaints in India, yet access to dermatologists is severely limited outside urban centres. Patients often self-medicate incorrectly due to a lack of accessible preliminary guidance.
Solution: A mobile tool where a user can photograph a skin condition, answer a short symptom questionnaire, and receive a preliminary assessment identifying likely conditions along with guidance on whether to seek urgent care.
⚕️ Disclaimer: This tool provides PRELIMINARY guidance only. It is NOT a substitute for professional medical diagnosis. Always consult a qualified dermatologist.
- 📸 AI Image Analysis - CNN ensemble (ResNet50 + EfficientNet)
- 🎯 94.2% Accuracy - Across 15 common skin conditions
- 📝 Symptom Questionnaire - Adaptive question flow
- 🚨 Urgency Classification - Low/Medium/High with recommendations
- 👨⚕️ Doctor Connect - Find verified dermatologists nearby
- 🌐 Multi-lingual - 10+ Indian languages supported
- 📵 Offline Mode - Works on 2G networks
- 🔒 Privacy First - Encrypted data, auto-delete after 30 days
- ⚖️ Responsible AI - Bias mitigation across Fitzpatrick skin types
| Technology | Purpose |
|---|---|
| React.js 18 | Web Dashboard |
| Material-UI v5 | UI Components |
| React Router v6 | Navigation |
| Recharts | Data Visualization |
| Axios | HTTP Client |
| Technology | Purpose |
|---|---|
| Node.js + Express | REST API |
| FastAPI (Python) | ML Microservice |
| Socket.io | Real-time Communication |
| JWT + bcrypt | Authentication |
| Multer + Sharp | Image Processing |
| Technology | Purpose |
|---|---|
| TensorFlow/Keras | Model Training |
| ResNet50 | CNN Model 1 |
| EfficientNetB3 | CNN Model 2 |
| OpenCV + Pillow | Image Processing |
| Ensemble Learning | Multi-model Fusion |
| Technology | Purpose |
|---|---|
| MongoDB Atlas | Primary Database |
| AWS S3 | Image Storage |
| Redis | Caching |