StitchMe

AI-powered wound assessment and treatment device with cross-platform applications for the University of Texas at San Antonio Senior Design capstone project. Transforms traditional wound care into an intelligent, automated system that combines computer vision, LiDAR scanning, and telemedicine capabilities.

🎯 Project Purpose

StitchMe revolutionizes wound care by creating an intelligent medical device that:

• AI-Powered Assessment: Computer vision analyzes wound severity and recommends treatment
• Cross-Platform Control: Flutter apps for iOS, Android, Web, Windows, macOS, and Linux
• LiDAR 3D Scanning: iPhone LiDAR creates precise wound measurements and 3D models
• Telemedicine Ready: Video calling connects patients with healthcare professionals
• Device Integration: Bluetooth/WiFi pairing between mobile apps and treatment device
• HIPAA Compliant: Secure data handling for medical information

🚀 Quick Start

# Clone and setup
git clone <repository-url>
cd StitchMe

# Run setup script
chmod +x scripts/setup-dev.sh
./scripts/setup-dev.sh

# Start Flutter app
cd apps/flutter_app && flutter run -d chrome
# Visit http://localhost:8080

🏗️ Architecture

StitchMe/
├── apps/
│   ├── flutter_app/         # Flutter app (Mobile + Desktop + Web)
│   ├── device/              # Native device app (Raspberry Pi/Jetson Nano)
│   │   ├── python/          # Main device controller application
│   │   └── arduino/         # Arduino sensor integration
│   └── native_modules/      # iOS LiDAR & Bluetooth integration
├── services/
│   ├── api/                 # Node.js + Express API server
│   ├── ai-service/          # Python + FastAPI AI microservice
│   └── realtime/            # WebRTC signaling server
├── packages/
│   ├── shared_models/       # Dart data models
│   ├── core_utils/          # Utilities & validators
│   ├── ui_kit/              # UI components & themes
│   └── device_sdk/          # Device communication SDK
└── infrastructure/          # Docker, deployment configs

✅ Development Status

Phase 1 — Foundation ✅ COMPLETE

• Monorepo structure with apps/, services/, and packages/
• Flutter cross-platform app with authentication UI
• Backend API structure with Node.js and Python services
• Development environment and build scripts ready

Phase 2 — Core Features 🔄 IN PROGRESS

• Camera integration and image processing
• iOS LiDAR 3D scanning with ARKit
• Basic AI wound detection and analysis
• Supabase backend integration

Phase 3 — AI & Communication ⏳ PLANNED

• Advanced wound classification models (PyTorch)
• WebRTC video calling system
• Device pairing and control protocols
• HIPAA compliance and security audit

🛠️ Technology Stack & Why

Our Choice: Flutter + Node.js + Python Hybrid

Why This Combination?

• Flutter: Single codebase for iOS, Android, Web, Desktop + LiDAR support
• Node.js API: Fast development + excellent WebRTC support for telemedicine
• Python AI Service: PyTorch and OpenCV for advanced computer vision and ML
• Supabase: Managed PostgreSQL + auth + real-time + HIPAA compliance options

Frontend Framework Comparison

Framework	Platforms	LiDAR	Performance	Medical Device Fit
Flutter ✅	iOS, Android, Web, Desktop	✅ iOS ARKit	⭐⭐⭐⭐⭐	✅ Perfect
React Native	iOS, Android, (Web)	✅ iOS ARKit	⭐⭐⭐⭐	⚠️ Limited desktop
Native Apps	Platform-specific	✅ Full access	⭐⭐⭐⭐⭐	❌ Too complex
Web Only	Web, PWA	❌ No access	⭐⭐⭐	❌ Hardware limitations

Backend Technology Comparison

Technology	Strengths	Medical Use Case	Team Fit
Node.js + Express ✅	WebRTC support, fast development	✅ Perfect for APIs	✅ Easy learning
Python + FastAPI ✅	AI/ML ecosystem, automatic docs	✅ Perfect for AI	✅ Great for ML
Java Spring Boot	Enterprise-grade, robust	⚠️ Overkill for MVP	❌ Too complex
.NET Core	Strong typing, performance	⚠️ Microsoft lock-in	❌ Platform limitations

Database & Services Comparison

Service	Pros	Cons	Medical Device Fit
Supabase ✅	PostgreSQL, real-time, auth, HIPAA	Newer service	✅ Perfect for MVP
Firebase	Google ecosystem, real-time	NoSQL limitations	⚠️ Limited complex queries
AWS Amplify	Full AWS integration	Complex, expensive	❌ Overkill for students
Custom PostgreSQL	Full control	Infrastructure overhead	❌ Too much work

🎯 Key Features

Mobile Apps (iOS/Android)

• Wound scanning with camera + LiDAR integration
• Device pairing with animated Bluetooth/WiFi connection
• Real-time AI wound analysis and treatment recommendations
• Video calling with healthcare professionals

Desktop & Web Apps

• Healthcare provider dashboard and device management
• Multi-device control and patient monitoring
• Video consultations and medical record management
• Hospital integration and compliance reporting

Device Native App (Raspberry Pi/Jetson Nano)

• Physical device control (motors, pumps, actuators)
• Real-time sensor monitoring (cameras, vitals, temperature)
• Safety systems and emergency stops
• Device status display on built-in touchscreen
• Communication with mobile/desktop apps via Bluetooth/WiFi

AI & Hardware

• Computer vision wound classification
• 3D LiDAR wound measurement and mapping
• Automated treatment recommendations
• Real-time vital signs monitoring

🛠️ Development

# Flutter app
cd apps/flutter_app && flutter run -d chrome

# Backend services
cd services/api && npm run dev
cd services/ai-service && python main.py

# All services
docker-compose up -d

🎓 Academic Project

University of Texas at San Antonio - Senior Design Capstone

• Timeline: Fall 2024 - Spring 2025 (2 semesters)
• Team: 4-5 Electrical and Computer Engineering students
• Objective: Design and build AI-powered medical device
• Deliverables: Working prototype, documentation, clinical testing

📄 License

Licensed under the Apache License 2.0 - see LICENSE file for details.

---

Ready for clinical testing 🏥 | AI-powered diagnostics 🤖 | Cross-platform deployment 📱💻🌐