Tuesday, June 13, 2023

Transforming Dermatology

Artificial intelligence has witnessed a rapid surge in the adoption of language models, with transformer-based pretrained language models (T-PLMs) playing a prominent role. These T-PLMs, such as BioBERT, MEP & Bioformer (BERT), Chatdoctor (LLaMA), OPAL & MedDialogue (BART), BioGPT and MedGPT (GPT 2), have revolutionized natural language processing in the biomedical research community, offering tailored performance for specific applications. Among these models, ChatGPT stands out due to its versatility, flexibility, and accessibility, making it widely applicable across various domains, including dermatology.

The fact that ChatGPT may be used by everyone, as opposed to models that can only be used by specialized developers and researchers, is one of its main advantages. This accessibility has paved the way for new possibilities in the field. By integrating clinical knowledge with interactive conversation, new generation of language models has the potential to transform all biomedical fields.

A recently posted preprint, open to peer review till July 8, focuses on the utilization of ChatGPT in dermatology. However, the study's conclusions have implications beyond dermatology and are relevant to other biomedical disciplines as well. Unlike previous reviews, this study goes further than performance evaluation by analyzing the actual utilization and real-world applications of ChatGPT in dermatology-related areas. By examining its practical use, the study offers valuable insights into the potential impact of ChatGPT in the broader biomedical research community. 


Gabashvili IS. ChatGPT in Dermatology: A Comprehensive Systematic Review JMIR Preprints. 02/06/2023:48979. medRxiv 2023.06.11.23291252; https://doi.org/10.1101/2023.06.11.23291252 

blockquote { margin:1em 20px; background: #dfdfdf; padding: 8px 8px 8px 8px; font-style: italic; }