Benny-Tang commited on
Commit
766411f
Β·
verified Β·
1 Parent(s): 0e5addd

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +10 -21
README.md CHANGED
@@ -9,32 +9,21 @@ app_file: app.py
9
  pinned: false
10
  ---
11
 
12
- # πŸ₯ AI Diagnostics Agent: Early Cancer Discovery
13
- AI-powered demo for early cancer risk detection using chest X-rays, lab markers, and explainable summaries.
14
 
15
- ⚠️ **Research demo only. Not for clinical use.**
16
-
17
- ## What it does
18
- This project demonstrates a multi-agent AI approach for early cancer risk detection:
19
- - **Imaging Agent** πŸ–ΌοΈ: Analyzes chest X-rays for findings relevant to lung cancer (*Mass, Nodule, Lung Opacity*).
20
- - Produces probability scores.
21
- - Display an optional **heatmap overlay** showing suspicious regions.
22
- - **Lab Agent** πŸ§ͺ: Parses tumor marker labs (PSA, CA125, AFP) from text and flags elevated results.
23
- - **Coordinator Agent** πŸ“‹: Combines imaging + lab results into an early cancer screening summary with disclaimers.
24
- - **Extra Samples**: MRI and CT text reports are included as input examples for future expansion.
25
 
26
  ---
27
 
28
- ## Try it online
29
- This app comes with demo data:
 
 
 
30
 
31
- - `samples/sample_xray1.png` β€” normal chest X-ray
32
- - `samples/sample_xray2.png` β€” suspicious chest X-ray
33
- - `samples/sample_labs.txt` β€” tumor marker results
34
- - `samples/sample_mri.txt` β€” example MRI report
35
- - `samples/sample_ct.txt` β€” example CT scan report
36
 
37
- ⚠️ Please **do not use real patient data**. This demo is for educational purposes only.
 
38
 
39
- ---
40
 
 
9
  pinned: false
10
  ---
11
 
12
+ # πŸ₯ AI Diagnostics Agent: Early Cancer Discovery (Demo)
 
13
 
14
+ **⚠️ Research demo only β€” not for clinical use**
 
 
 
 
 
 
 
 
 
15
 
16
  ---
17
 
18
+ ### πŸš€ Overview
19
+ AI Diagnostics Agent is a **proof-of-concept early cancer screening tool** combining:
20
+ - **Imaging Agent** β€” analyzes chest X-rays using TorchXRayVision to detect lung anomalies (mass, nodule, opacity) and generate heatmaps.
21
+ - **Lab Agent** β€” parses tumor marker results (PSA, CA125, AFP) for abnormal readings.
22
+ - **AI Coordinator (Moonshot.ai)** β€” uses a large language model to summarize overall cancer risk, highlight abnormalities, and suggest next steps in human-readable form.
23
 
24
+ ---
 
 
 
 
25
 
26
+ ### 🧠 Powered by Moonshot.ai
27
+ The AI Coordinator uses [Moonshot.ai](https://www.moonshot.cn) LLM API to generate intelligent summaries.
28
 
 
29