Google Drive Link for video demonstration - https://drive.google.com/drive/folders/1Vwj1798ZISMxuLQVPyvjWrD14ump6-RL?usp=sharing
Early Detection Engine for Silent Diseases
A large percentage of life-threatening diseases—including diabetes, hypertension, liver disorders, mental health conditions, and early cardiac risk—remain undiagnosed for years because symptoms are mild, ignored, or fragmented across reports. Healthcare systems today rely heavily on reactive diagnosis rather than predictive prevention, leading to late detection, higher treatment costs, and worse patient outcomes.
Our Challenge: Build an AI-driven early risk detection system that:
- Aggregates non-obvious health signals (lab trends, lifestyle data, mental health indicators, family history)
- Detects silent or early-stage diseases before clinical diagnosis
- Generates risk probability scores, not binary outcomes
- Provides preventive action recommendations for both patients and doctors
Four Core Modules:
-
📄 LabReport Analyzer
Upload your lab reports (PDF/images) → AI extracts trends and flags anomalies over time. -
🤖 Signal AI Chatbot
A conversational health assistant that discusses symptoms, provides guidance, and answers medical questions. -
📋 ImpChoice Health Assessment
MCQ-based evaluation of family history + current lifestyle → generates personalized SilentScore and risk breakdowns. -
📊 Health Dashboard
Weekly scorecards, health missions, doctor tools, and progress tracking—all in one unified interface.
Frontend: React + Tailwind CSS + Vercel AI SDK
Backend: FastAPI (Python) + TypeScript/JavaScript
Database: PostgreSQL
AI/ML: Hugging Face + GTWY.ai
Architecture: Microservices, REST APIs, JWT Authentication- Multi-Year Trend Analysis – Not just single reports, but health trajectory over time
- Probability-Based Risk Scores – No binary "yes/no," just intelligent likelihood percentages
- Holistic Data Fusion – Lab + lifestyle + mental health + family history in one profile
- Prevention-Focused Missions – Gamified, actionable health tasks to reduce risks
- Doctor & Patient Tools – Designed for both ends of the care continuum
# Clone the repository
git clone https://github.com/HelloWorld/silentsignal-ai.git
cd silentsignal-ai
# Backend Setup
cd backend
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
pip install -r requirements.txt
# Set up environment variables
cp .env.example .env
# Add your API keys to .env (HuggingFace, Vellum AI, GTWY.ai, etc.)
# Run backend
uvicorn main:app --reload --port 8000
# Frontend Setup (new terminal)
cd ../frontend
npm install
# Set up frontend environment
cp .env.example .env.local
# Configure REACT_APP_API_URL=http://localhost:8000
# Run frontend
npm startAccess the application: http://localhost:3000
✅ First platform to combine:
- Longitudinal lab data analysis
- Real-time lifestyle tracking
- Family history pattern recognition
- Mental health indicators
✅ Medical-grade chatbot with:
- Fine-tuned on symptom-disease datasets
- Context-aware conversations
- Emergency red flag detection
✅ Explainable AI scoring:
- Each risk factor is traceable
- Confidence intervals provided
- Medical guideline alignment| Area | Impact |
|---|---|
| Early Detection | Identify silent diseases 3-5 years earlier |
| Cost Reduction | Preventative care lowers treatment costs by 60% |
| Doctor Efficiency | Saves 20+ minutes per consultation |
| Patient Empowerment | Transforms passive patients into proactive managers |
SilentSignal AI – Because the best treatment is the one that never has to happen.
"Detect what patients don't even know they have."