D

Deid Bert I2b2

Developed by obi
This model is used to identify and remove protected health information (PHI/PII) from medical records in compliance with HIPAA privacy standards.
Downloads 129.39k
Release Time : 3/2/2022

Model Overview

A sequence labeling model fine-tuned on ClinicalBERT, specifically designed for de-identification processing of electronic health records (EHR), capable of recognizing 11 categories of protected health information entities.

Model Features

HIPAA Compliance
Supports identification of 11 categories of protected health information (PHI) entities as defined by HIPAA
Clinical Context Optimization
Fine-tuned on Bio_ClinicalBERT, specifically optimized for medical text characteristics
Context Awareness
Utilizes a 32-token context window to enhance sentence boundary recognition

Model Capabilities

Medical Entity Recognition
Sensitive Information Detection
Text De-identification Processing
Sequence Labeling Prediction

Use Cases

Medical Data Privacy Protection
Electronic Health Record Anonymization
Automatically removes patient personal information before sharing medical records
See performance section for metrics like F1 score
Clinical Research Data Preparation
Cleans sensitive information from medical records for research purposes
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase