Large Language Models and Otolaryngology: A Review | JoVE Visualize

Area of Science:

Artificial intelligence
Natural language processing
Medical informatics

Background:

Large language models (LLMs) are advancing rapidly, with significant potential to revolutionize healthcare administration, clinical practice, and research.
While AI adoption is accelerating across medicine, otolaryngology (ENT) has lagged behind other specialties in leveraging LLMs.
Otolaryngology's reliance on complex, multimodal data (text, imaging, electrophysiology, video) presents unique opportunities for LLM application.

Purpose of the Study:

To review the advantages and innovations of LLMs relevant to otolaryngology.
To provide otolaryngologists with a foundation for understanding and advancing LLM technologies within their field.
To encourage the broader adoption and application of LLMs in otolaryngology research and patient care.

Main Methods:

Review of existing literature on LLM applications in medicine, with a focus on innovations in other specialties.
Analysis of otolaryngology's specific data characteristics and clinical complexities in relation to LLM capabilities.
Identification of current limitations and future directions for LLM implementation in otolaryngology.

Main Results:

LLMs offer powerful language abilities by combining deep learning and natural language processing.
Other medical specialties have successfully applied LLMs to structured data conversion, automated phenotyping, administrative task streamlining, decision support, and multimodal data integration.
Otolaryngology research has primarily focused on limited question-answering tasks, often using closed-source models, hindering clinical utility.

Conclusions:

LLMs hold significant promise for enhancing otolaryngology research and patient care due to the field's rich data.
Moving beyond feasibility studies to clinical validation, open-source development, and domain-specific fine-tuning is crucial for progress.
Responsible LLM implementation, including secure deployment and legal oversight, can improve efficiency, decision support, and patient outcomes in otolaryngology.