deid roberta i2b2

提供商obi
分类token-classification
许可证mit
下载量470.4K
星标0

简介

deid-roberta-i2b2 是一款专注于医疗健康领域脱敏（De-identification）的命名实体识别模型。它基于 RoBERTa 架构，并在权威的 i2b2 医疗数据集上进行了微调，能够精准识别医疗记录中的患者姓名、日期、地点等隐私信息。对于需要处理电子病历（EHR）或临床文本的中国开发者来说，该模型提供了高效的自动化脱敏方案，上手难度低，可直接集成到医疗数据预处理流水线中，替代繁琐的正则表达式，显著提升数据隐私保护的自动化程度。

核心亮点

  • 专为医疗文本设计，精准识别隐私实体
  • 基于 RoBERTa 架构，具备强语义理解能力
  • 适配 i2b2 标准，医疗脱敏场景开箱即用
  • MIT 协议开源，方便企业级商业集成

使用方法

安装依赖
# 安装 Hugging Face transformers
pip install transformers torch
SDK 使用
# 使用 transformers 加载模型
from transformers import AutoModel, AutoTokenizer

model = AutoModel.from_pretrained("obi/deid_roberta_i2b2")
tokenizer = AutoTokenizer.from_pretrained("obi/deid_roberta_i2b2")

Hugging Face 下载

我们推荐使用命令行或者 Hugging Face Hub SDK 来进行模型的下载。

操作指引：在下载前，请先通过如下命令安装 huggingface_hub：

操作指引 
pip install -U huggingface_hub

命令行下载

下载完整模型库

下载完整模型库 
huggingface-cli download obi/deid_roberta_i2b2

下载单个文件到指定本地文件夹（以下载 config.json 到当前路径下 ./dir 目录为例）

下载单个文件到指定本地文件夹（以下载 config.json 到当前路径下 ./dir 目录为例） 
huggingface-cli download obi/deid_roberta_i2b2 config.json --local-dir ./dir

更多命令行下载选项，可参见官方文档

SDK 下载

SDK 下载 
# 模型下载
from huggingface_hub import snapshot_download
model_dir = snapshot_download('obi/deid_roberta_i2b2')

Git 下载

请确保 lfs 已经被正确安装

Git 下载 
git lfs install
git clone https://huggingface.co/obi/deid_roberta_i2b2

如果您希望跳过 lfs 大文件下载，可以使用如下命令

跳过 LFS 
GIT_LFS_SKIP_SMUDGE=1 git clone https://huggingface.co/obi/deid_roberta_i2b2

模型文件托管在 Hugging Face Hub，使用 HF CLI / SDK / Git 直接下载，不经过本站。

PyTorch / Transformers 使用

安装 Transformers

安装 Transformers 
pip install -U transformers torch

模型加载和推理

模型加载和推理 
from transformers import AutoModelForCausalLM, AutoTokenizer

model = AutoModelForCausalLM.from_pretrained('obi/deid_roberta_i2b2')
tokenizer = AutoTokenizer.from_pretrained('obi/deid_roberta_i2b2')

完整文档

来源: HuggingFace

---
language:
- en
thumbnail: "https://www.onebraveidea.org/wp-content/uploads/2019/07/OBI-Logo-Website.png"
tags:

  • deidentification

  • medical notes

  • ehr

  • phi

datasets:
  • I2B2

metrics:
  • F1

  • Recall

  • Precision

widget:
  • text: "Physician Discharge Summary Admit date: 10/12/1982 Discharge date: 10/22/1982 Patient Information Jack Reacher, 54 y.o. male (DOB = 1/21/1928)."

  • text: "Home Address: 123 Park Drive, San Diego, CA, 03245. Home Phone: 202-555-0199 (home)."

  • text: "Hospital Care Team Service: Orthopedics Inpatient Attending: Roger C Kelly, MD Attending phys phone: (634)743-5135 Discharge Unit: HCS843 Primary Care Physician: Hassan V Kim, MD 512-832-5025."

license: mit
---

Model Description

  • A RoBERTa [[Liu et al., 2019]](https://arxiv.org/pdf/1907.11692.pdf) model fine-tuned for de-identification of medical notes.
  • Sequence Labeling (token classification): The model was trained to predict protected health information (PHI/PII) entities (spans). A list of protected health information categories is given by HIPAA.
  • A token can either be classified as non-PHI or as one of the 11 PHI types. Token predictions are aggregated to spans by making use of BILOU tagging.
  • More details on how to use this model, the format of data and other useful information is present in the GitHub repo: Robust DeID.

How to use

  • Steps on how this model can be used to run a forward pass can be found here: Forward Pass
  • In brief, the steps are:
* Sentencize (the model aggregates the sentences back to the note level) and tokenize the dataset. * Use the predict function of this model to gather the predictions (i.e., predictions for each token). * Additionally, the model predictions can be used to remove PHI from the original note/text.

Dataset

  • The I2B2 2014 [[Stubbs and Uzuner, 2015]](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4978170/) dataset was used to train this model.

| | I2B2 | | I2B2 | |
| --------- | --------------------- | ---------- | -------------------- | ---------- |
| | TRAIN SET - 790 NOTES | | TEST SET - 514 NOTES | |
| PHI LABEL | COUNT | PERCENTAGE | COUNT | PERCENTAGE |
| DATE | 7502 | 43.69 | 4980 | 44.14 |
| STAFF | 3149 | 18.34 | 2004 | 17.76 |
| HOSP | 1437 | 8.37 | 875 | 7.76 |
| AGE | 1233 | 7.18 | 764 | 6.77 |
| LOC | 1206 | 7.02 | 856 | 7.59 |
| PATIENT | 1316 | 7.66 | 879 | 7.79 |
| PHONE | 317 | 1.85 | 217 | 1.92 |
| ID | 881 | 5.13 | 625 | 5.54 |
| PATORG | 124 | 0.72 | 82 | 0.73 |
| EMAIL | 4 | 0.02 | 1 | 0.01 |
| OTHERPHI | 2 | 0.01 | 0 | 0 |
| TOTAL | 17171 | 100 | 11283 | 100 |

Training procedure

  • Steps on how this model was trained can be found here: Training. The "model_name_or_path" was set to: "roberta-large".
* The dataset was sentencized with the en_core_sci_sm sentencizer from spacy. * The dataset was then tokenized with a custom tokenizer built on top of the en_core_sci_sm tokenizer from spacy. * For each sentence we added 32 tokens on the left (from previous sentences) and 32 tokens on the right (from the next sentences). * The added tokens are not used for learning - i.e, the loss is not computed on these tokens - they are used as additional context. * Each sequence contained a maximum of 128 tokens (including the 32 tokens added on). Longer sequences were split. * The sentencized and tokenized dataset with the token level labels based on the BILOU notation was used to train the model. * The model is fine-tuned from a pre-trained RoBERTa model.
  • Training details:
* Input sequence length: 128 * Batch size: 32 (16 with 2 gradient accumulation steps) * Optimizer: AdamW * Learning rate: 5e-5 * Dropout: 0.1

Results

Questions?

Post a Github issue on the repo: Robust DeID.

查看官方来源