NeuraLens

• 

  Dashbaord View

• 

  Mutiple Model Assessment

• 

  Home Page

Inspiration

Neurological disorders often go undetected until it’s too late. We wanted to build something that empowers early screening using AI and accessible tools.

What it does

NeuraLens integrates speech analysis, retinal imaging, motor assessment, and cognitive testing into one AI-powered platform for early risk detection.

How we built it

Frontend: Next.js + TailwindCSS (PWA ready) Backend: FastAPI + PostgreSQL + Redis ML: TensorFlow/PyTorch + XGBoost + OpenCV + Librosa Infra: Dockerized microservices with synthetic demo data

Challenges we ran into

Aligning multiple ML pipelines into a single scoring framework Generating usable synthetic datasets for the demo Achieving real-time analysis speed without losing accuracy Ensuring frontend ↔ backend integration worked in limited hackathon time

Accomplishments that we're proud of

Created a working prototype that showcases multi-modal AI fusion Designed a clear risk scoring pipeline (NRI Fusion) Built modular architecture that can scale beyond hackathon scope

What we learned

Integrating multiple modalities is harder than training models in isolation Hackathon MVP ≠ full product, but clarity in architecture helps pivot fast User experience matters as much as backend accuracy

What's next for NeuraLens

Expand motor function analysis (tremor quantification) Clinical validation with real datasets Full PWA deployment with offline support Partnerships with health organizations for global accessibility

Built With

• chartjs
• docker-apis/tools:-librosa
• fastapi
• next.js
• nextjs
• opencv
• postgresssql
• python
• pytorch
• react
• redis
• scikit-learn
• scikit-learn-databases/infra:-postgresql
• tailwainfcss
• tailwindcss-ml/ai:-tensorflow
• tensorflow
• typescript
• typescript-frameworks:-fastapi
• xgboost