Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Jun 23, 2026

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

Explainable AI for mental health emergency returns: integrating large language models with predictive modeling.

Abdulaziz Ahmed1,2, Mohammad Saleem1, Mohammed Alzeen1

  • 1Department of Health Services Administration, School of Health Professions, University of Alabama at Birmingham, Birmingham, AL 35233, United States.

JAMIA Open
|June 22, 2026
PubMed
Summary

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Exploration of molecular mechanisms responsible for arthritis alleviating attributes of Calliandra haematocephala extracts.

Inflammopharmacology·2026
Same author

A sodium-HIF1α axis coordinates immune metabolic reprogramming and mitochondrial remodeling in salt-sensitive hypertension.

Research square·2026
Same author

Periodontal Health Status, Awareness, and Oral Hygiene Practices among Fixed and Removable Partial Denture Wearers in Al-Ahsa, Saudi Arabia.

Journal of International Society of Preventive & Community Dentistry·2026
Same author

Effect of methanolic extract of Salix babylonica on biochemical parameters and gene expression in streptozotocin induced diabetic rat model.

Fitoterapia·2026
Same author

Characterizing dietary diversity among urban office workers in northern Ghana: a cross-sectional study.

Journal of health, population, and nutrition·2026
Same author

A one-week reduced-carbohydrate diet lowers insulin requirements and shifts the IGF axis with no detectable short-term change in endothelial function in a randomized, crossover trial of adults with type 1 diabetes.

Cardiovascular diabetology. Endocrinology reports·2026

Large Language Models (LLMs) integrated with machine learning (ML) modestly improved predicting emergency department (ED) returns for mental health (MH) patients. This hybrid approach significantly enhanced model interpretability, offering better clinical decision support.

Area of Science:

  • Artificial Intelligence in Healthcare
  • Clinical Decision Support Systems
  • Mental Health Informatics

Background:

  • Emergency department (ED) returns for mental health (MH) conditions represent a significant healthcare burden.
  • Traditional machine learning (ML) models show potential for predicting these returns but lack clinical interpretability.
  • Enhancing the explainability of predictive models is crucial for clinical adoption.

Purpose of the Study:

  • To evaluate the integration of Large Language Models (LLMs) with ML methods for predicting 30-day ED returns in MH patients.
  • To assess whether this hybrid approach can improve both predictive performance and model interpretability.
  • To explore the utility of LLM-derived features and a novel LLM-SHAP framework for enhanced explanations.

Main Methods:

Keywords:
30 days emergency returnEmergency departmentExplainable AIlarge language modelmachine learning

Related Experiment Videos

Last Updated: Jun 23, 2026

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
03:14

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

  • Retrospective analysis of 42,464 ED visits for 27,904 unique MH patients (2018-2022).
  • LLaMA 3 (8B) used for few-shot classification of chief complaints and social determinants of health (SDoH).
  • LLM-extracted features integrated into an XGBoost model; LLM-SHAP framework for enhanced interpretability.

Main Results:

  • LLM chief complaint classifier achieved high accuracy (0.882) and recall (0.88).
  • SDoH classifiers demonstrated strong performance (F1-scores 0.67-0.96).
  • Hybrid model improved predictive performance (AUC from 0.74 to 0.76) and significantly enhanced interpretability via LLM-SHAP.

Conclusions:

  • Integrating LLMs with ML modestly improves predictive accuracy for ED returns in MH patients.
  • This hybrid approach substantially enhances model interpretability, providing contextualized risk explanations.
  • The findings suggest a promising pathway for developing actionable and explainable clinical decision support tools in emergency psychiatry.