Skip to content

AlanRoybal/HackDartmouth2025

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

46 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Neurolytics – AI‑Powered Neuro‑Oncology Assistant

Neurolytics Landing


Overview

More than 50 million people worldwide need MRI interpretation every year, yet neuroradiologists are scarce and expensive. Neurolytics lets anyone upload a brain MRI, receive an AI‑driven diagnostic summary, and ask follow‑up questions in natural language – all inside a single web page.

We combine Next.js, Google Gemini Vision, and AWS S3 to deliver expert‑level insights in seconds, freeing clinicians to focus on critical cases and empowering students to learn neuro‑oncology hands‑on.


What It Does

Step Action Tech
1 Upload any T1/T2 slice (JPEG/PNG) Next.js file input
2 Instant Analysis of tumor type, grey‑matter loss, abnormalities Gemini Vision 2.0
3 One‑Click Chat – ask anything about the scan with grounded answers Gemini LLM + JSON context
4 Auto‑Archival – each run saved to S3 and visible in History tab AWS S3

Key Innovations

  • Context‑Grounded Chat – every reply cites the exact JSON analysis, avoiding hallucinations.
  • Timestamp Keys – frontend & backend stay perfectly in‑sync even with concurrent uploads.
  • Zero‑Sound UX – purely visual + textual; ideal for quiet clinical settings.
  • Plug‑and‑Play Cloud – no GPUs needed; Gemini handles all computation.

System Architecture

User ─▶ Next.js ──▶ Flask API ──▶ Gemini Vision
        │  ▲          │
        │  │          └──▶ AWS S3  (images + JSON + summary)
        ▼  │
    Chat UI ◀──────────┘
  1. /analyze_mri → uploads image, gets JSON analysis, stores to saved/<timestamp>/.
  2. /chat → receives { prompt, timestamp }, loads matching JSON, returns answer.
  3. /history → lists past runs with thumbnails & summaries.

Built With

  • FrontendNext.js 14, TypeScript, TailwindCSS, shadcn/ui
  • BackendPython 3.10, Flask, google-generativeai, boto3
  • CloudAWS S3, GitHub Actions (CI/CD)

API Reference

Verb Endpoint Payload Returns
POST /analyze_mri form‑data file { timestamp, image_url, … }
POST /chat { prompt, timestamp } { response }
GET /history [{ timestamp, mri_url, summary }]

Future Work

  • Segmentation overlay masks
  • DICOM upload + series handling
  • HIPAA‑grade audit logging

Check Us Out

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •