Advancing Mental Health Research through AI & Clinical Data Modernization

Advancing Mental Health Research through AI & Clinical Data Modernization

Wednesday, March 5, 2025 12:00 PM to 1:00 PM · 1 hr. (US/Pacific)
Venetian | Level 1 | Casanova 501
Workforce ConneXtions
Data and Information

Information

Note: This poster will be on display at Workforce ConneXtions on Wednesday, March 5, 2025 from 10 am - 4 pm. Meet the author(s) from 12 pm - 1 pm.


This project leverages real-world data from Electronic Health Records (EHR) to improve psychiatric care by overcoming the challenges of unstructured and isolated data. By developing a Clinical Data Mart focused on psychiatric data within the UTHealth system, the project aims to enhance the understanding and treatment of mental health. It includes three key aims: (1) Data modernization through standardization and integration, (2) the development of Psychiatric Large Language Models (LLMs) for data analysis, and (3) participation in case studies to explore psychosocial predictors like suicide. This work supports a learning behavioral health system for more effective mental health care.

Sub-Topic Category
Artificial Intelligence/Machine Learning
Target Audience
Clinical InformaticistData ScientistEarly Careerist
Level
Introductory
CEU Type
CAHIMSCPDHTSCPHIMS
Format
60-Minute Case Study
Learning Objective #1
Analyze the challenges posed by unstructured psychiatric data in EHR systems and propose methods to overcome them using data standardization techniques
Learning Objective #2
Create a novel Clinical Data Mart by integrating historical and current psychiatric data, enabling advanced AI models for predictive analytics in mental health research
Learning Objective #3
Evaluate the effectiveness of Large Language Models (LLMs) like BERT and GPT in identifying clinical phenotypes from psychiatric notes, considering their impact on diagnostic accuracy and treatment outcomes
Learning Objective #4
Apply AI and NLP techniques to develop tools that identify psychosocial predictors such as stressful life events and suicidality from psychiatric notes
Learning Objective #5
Synthesize findings from observational case studies to refine and validate the Clinical Data Mart's application in real-world behavioral health systems
Session #
WFC-19