metadata

title: MedSynapticGPT
emoji: 🩺
colorFrom: red
colorTo: blue
sdk: streamlit
sdk_version: 1.35.0
app_file: app.py
pinned: false

MedSynapticGPT – Multimodal Clinical Reasoner

Tagline: “Read the scan. Hear the patient. Answer like an expert.”

MedSynapticGPT brings the creative power of GPT-4o to clinical imaging, notes, and speech. Upload a chest X-ray (DICOM/PNG), paste a discharge summary, or record symptoms—get structured impressions with SNOMED-CT codes, guideline citations, and treatment suggestions.

🚑 Core Modules

Tab	What it does	Key APIs
Radiology AI	Vision interpretation, abnormality detection, TNM staging	`gpt-4o vision`, `pydicom`
Clinical Note Q&A	Summarize or answer free-form questions	`gpt-4o`
Voice Triage	Transcribe symptoms, suggest differential	`whisper-1`, `gpt-4o`
UMLS Lookup	Search concepts, synonyms, codes	UMLS REST + caching
GraphRAG Explorer	Prototype biomedical KG Q&A	`networkx`, toy graph

Quickstart

git clone <repo>
cd MedSynapticGPT
pip install -r requirements.txt
export OPENAI_API_KEY="sk-…"
streamlit run app.py

Deploy on Spaces → add secrets. For on-prem, launch uvicorn api:app.

Monetization

Freemium 3 studies/day → Stripe pay-per-usage.
Pro ($49/mo) – unlimited, PDF reports, HL7 export.
Enterprise API – hospital integration, HIPAA BAA.