基本信息

模态

文本

大小

0.008675 GB

许可

MIT

创建/更新时间

2026-10-08

版本

main

资源简介

MedChat诊断报告

【数据集背景】

本数据集源站标题为《Purdue-M2/MedChat》。源站首次发布：2025-06-02。

【数据内容】

源站原始描述：MedChat > Philip Liu, Sparsh Bansal, Jimmy Dinh, Aditya Pawar, Ramani Satishkumar, Shail Desai, Neeraj Gupta, Xin Wang, Shu Hu A modular framework that uses modern vision back-ends with role-specialised LLM agents to draft glaucoma diagnostic reports from retinal fundus images. The core idea is for each agent to focus on a narrow clinical role, while a *Director* agent synthesises their opinions into a concise, clinically-grounded report. - Method Overview 1. Vision Pre-processing * Classifier…

【适用方向】

文本类数据可用于医学自然语言处理，如命名实体识别、信息抽取与文本分类。

【获取与许可】

源站页面：https://github.com/Purdue-M2/MedChat

使用许可：MIT。

【使用注意】

· 源站最后修改于 2025-06-08，引用时请注意版本时效。

原始链接

https://github.com/Purdue-M2/MedChat

访问原始数据

官方服务

专家保障全链路合规授权与标注交付

陕川双基地 × 超十家三甲医院直连

从合规授权、采集治理、专业标注到数据集交付，我们提供全链路闭环服务，让高质量临床数据即拿即用。

⚡️ 需要原始数据或标注支持？联系我们获取定制方案。

帮我联系

下载信息

注册下载

需要注册 Kaggle 账号并登录后下载，适合需要跟踪下载记录和使用 API 的用户。

暂未开放

公开下载

无需注册即可直接获取公开样本或文档，适合快速预览和评估数据集质量。

免登录

有偿下载

公开数据集受托下载与技术交付服务。

提供高速下载与技术交付服务（收技术服务费，非数据销售）

暂未开放

当前数据集主要来源为 Kaggle 公开托管，完整影像包建议通过原始链接或 Kaggle API 获取。

使用方式

数据集获取

git clone https://github.com/Purdue-M2/MedChat.git

curl -L -o repo.zip https://github.com/Purdue-M2/MedChat/archive/refs/heads/main.zip
unzip repo.zip

源站 README 摘录（使用方式）

MedChat

Philip Liu, Sparsh Bansal, Jimmy Dinh, Aditya Pawar, Ramani Satishkumar, Shail Desai, Neeraj Gupta, Xin Wang, Shu Hu
A modular framework that uses modern vision back-ends with role-specialised LLM agents to draft glaucoma diagnostic reports from retinal fundus images. The core idea is for each agent to focus on a narrow clinical role, while a Director agent synthesises their opinions into a concise, clinically-grounded report.

Method Overview

Vision Pre-processing
- Classifier (SwinV2) → glaucoma probability p (binned into “no glaucoma / possible glaucoma / likely glaucoma / glaucoma detected”).
- Segmentor (SegFormer) → optic-cup & disc masks → cup-to-disc ratio (CDR).
Core Prompt
Natural-language sentences summarise p and CDR, optionally appended with clinician notes.
Role Generation
A meta-prompt asks GPT-4.1 to list relevant clinical roles (e.g., Ophthalmologist, Optometrist, Pharmacist).
Role-Specialised Sub-Reports
Each role gets the core prompt plus narrow instructions and writes a focused sub-report.
Director Synthesis
Another GPT-4.1 instance combines all sub-reports, resolves minor conflicts, and produces one clean diagnostic report.
Output & Interface
The final report, probability, CDR, and sub-reports are returned to the client (e.g., MedChat front-end) for interactive Q&A and PDF export.

✨ Key Contributions

#	Contribution	Why it matters
1	Multi-agent reasoning: Ophthalmologist, Optometrist, Pharmacist, … plus a Director agent	Reduces hallucinations and reflects real-world collaboration
2	Tight CAD ⇄ LLM loop: SwinV2 classifier (glaucoma probability) + SegFormer segmentor (optic-cup/optic-disc masks)	Keeps language output anchored to verifiable image features (e.g., CDR)
3	MedChat interface (browser-based)	Enables interactive Q&A and PDF report download for clinicians and learners – see `frontend/` for code; no external link included

Repository Layout

├── backend/              # Python API, multi-agent pipeline, model wrappers
│   ├── cad/              #  SwinV2 classifier &amp; SegFormer segmentor
│   ├── agents/           #  Role prompts, Director logic
│   └── api.py            #  FastAPI / Flask endpoints
├── frontend/             #


## 数据加载示例（表格/文本类）

```python
import pandas as pd, glob, os

files = (glob.glob(os.path.join(path, "**", "*.csv"), recursive=True)
       + glob.glob(os.path.join(path, "**", "*.tsv"), recursive=True)
       + glob.glob(os.path.join(path, "**", "*.xlsx"), recursive=True))
print("数据文件:", files)
df = pd.read_csv(files[0])
print(df.shape); print(df.columns.tolist()); print(df.head(3))