Benchmarking Modern Named Entity Recognition Techniques for Free-text Health Record De-identification
This work addresses the problem of making EHR data publicly available for research by improving de-identification techniques, but it is incremental as it benchmarks existing methods on a specific dataset.
The study evaluated deep learning-based named entity recognition methods for de-identifying electronic health records, finding that BiLSTM-CRF performed best, with character embeddings and CRFs improving precision but reducing recall, and transformers underperforming as context encoders.
Electronic Health Records (EHRs) have become the primary form of medical data-keeping across the United States. Federal law restricts the sharing of any EHR data that contains protected health information (PHI). De-identification, the process of identifying and removing all PHI, is crucial for making EHR data publicly available for scientific research. This project explores several deep learning-based named entity recognition (NER) methods to determine which method(s) perform better on the de-identification task. We trained and tested our models on the i2b2 training dataset, and qualitatively assessed their performance using EHR data collected from a local hospital. We found that 1) BiLSTM-CRF represents the best-performing encoder/decoder combination, 2) character-embeddings and CRFs tend to improve precision at the price of recall, and 3) transformers alone under-perform as context encoders. Future work focused on structuring medical text may improve the extraction of semantic and syntactic information for the purposes of EHR de-identification.