dlyog/synth_clin_notes

# Synthetic Clinical Notes Dataset

This dataset, generated using LLAMA2, is designed to mimic FHIR Document Reference Clinical Notes. It follows a layout similar to MIMIC, but it's important to note that this dataset contains no Personal Health Information (PHI) or Personally Identifiable Information (PII).

## Dataset Details

- **Name**: Synthetic Clinical Notes Dataset

## Key Features

- **Synthetic Data**: All data in this dataset is synthetic, ensuring no risk of exposing real patient information.
- **FHIR Document Reference Layout**: The dataset closely mirrors the structure and format of FHIR Document Reference Clinical Notes, making it suitable for healthcare-related machine learning tasks.
- **MIMIC-style Layout**: For researchers familiar with the MIMIC dataset, this synthetic dataset offers a similar layout, facilitating a smoother transition.

## Usage

The dataset is compatible with the Hugging Face Datasets library. Here's a quick start guide:

```python
from datasets import load_dataset

# Load the synthetic clinical notes dataset
dataset = load_dataset("your_huggingface_dataset_name_here")

# Exploring the dataset
print(dataset["train"][0]) # Print the first entry from the training set

原始信息汇总

合成临床笔记数据集

数据集详情

• 名称: 合成临床笔记数据集

关键特性

• 合成数据: 数据集中的所有数据均为合成数据，确保不会泄露真实患者信息。
• FHIR文档参考布局: 数据集模仿FHIR文档参考临床笔记的结构和格式，适用于医疗相关的机器学习任务。
• MIMIC风格布局: 对于熟悉MIMIC数据集的研究人员，该合成数据集提供类似的布局，便于更顺畅地过渡。

使用方法

数据集与Hugging Face Datasets库兼容。以下是一个快速入门指南：

python from datasets import load_dataset

加载合成临床笔记数据集

dataset = load_dataset("your_huggingface_dataset_name_here")

探索数据集

print(dataset["train"][0]) # 打印训练集中的第一个条目