… In this work, we developed a novel GPT4-enabled de-identification framework (“DeID-GPT”) …
used medical text data de-identification methods, our developed DeID-GPT showed the …

… of the de-identification process, surpassing the zero-shot … the performance of a zero-shot
learning GPT-4 model, with 99% … DeIDGPT demonstrates higher accuracy (0.99) and reliability …

… use of GPT-4 for de-identifying medical reports, reporting that GPT-4, using a zero-shot prompt…
within the GPT-based de-identification process, we sent the data to GPT-4 three times and …

… the de-identification task as a set of instructions for GPT-4, … , suggesting that zero-shot
named entity recognition is still a … DeID-GPT: Zero-shot Medical Text De-Identification by GPT-4…

… On March 14, 2023, Open AI released GPT-4, a significant advancement in large language
… use of GPT for de-identification of clinical notes. Currently, only synthetic medical data can be …

… models such as GPT-3.5 and GPT-4. Although the review comprehensively describes …
zero-shot learning, highlights the growing synergy between artificial intelligence and healthcare …

… DeID-GPT [31] is a framework enabled by GPT-4 that automatically identifies and removes
identifying information. The HIPAA guidelines and PHI identifiers are incorporated into the de…

… One potential avenue for enhancing GPT-4's consistency lies … , into API-driven interactions
with GPT-4, our findings suggest … Deid-gpt: zero-shot medical text de-identification by gpt-4. …

… of GPT-4 in patient eligibility classification for clinical trials. This research asks the following
research questions: Can GPT-4 … Deid-gpt: Zero-shot medical text de identification by gpt-4. …

… , while zeroshot LLaMA 2 outperforms others, GPT-4 still achieves a … zero-shot learning,
fine-tuning significantly enhances its performance, elevating it to a level comparable with GPT-4, …