Advanced Informatics for Mental Health Lab Research

We utilize advanced natural language processing (NLP) techniques and large language models (LLMs) such as GPT, Claude, and Llama to explore the potential of AI tools in enhancing the management of pediatric developmental, behavioral, and mental health conditions. Our research focuses on assessing the feasibility and accuracy of these AI technologies to improve quality of care by analyzing clinical notes, extracting valuable insights from unstructured free-text data, and evaluating quality of care. Additionally, we investigate how AI can interpret patient timelines, identify trends in care delivery, and uncover critical information that may not be easily captured through traditional methods. This research aims to develop AI-driven solutions that can assist healthcare providers in delivering more accurate, timely, and evidence-based care for children facing these conditions.

Methods include advanced quantitative techniques to analyze both structured and unstructured (free text) electronic health record (EHR) data. This includes biostatistical methods, advanced data analytics, applied machine learning models, and natural language processing (NLP).

Relevant Publications

Applying Large Language Models to Assess Quality of Care: Monitoring ADHD Medication Side Effects

Measuring quality-of-care in treatment of young children with attention-deficit/hyperactivity disorder using pre-trained language models

Using large-scale EHR databases, such as those from Stanford's community-based network Packard Children's Health Alliance (PCHA) and PEDSnet, our research focuses on conducting population health and health services research to identify patterns and trends in the delivery of care for pediatric developmental, behavioral, and mental health conditions across different populations. By analyzing this rich clinical data, we also aim to better understand the relationships between demographic, family, and other variables with clinical and quality of care outcomes. This analysis helps us pinpoint critical gaps in care and develop actionable recommendations to enhance the quality and accessibility of evidence-based health services for children facing these conditions.

Research methods include quantitative techniques, such as biostatistical methods and advanced data analytics, as well as qualitative approaches like interviews, focus groups, and surveys with primary care clinicians, patient families, and other key stakeholders.

Relevant Publications

Detection of Speech-Language Delay in the Primary Care Setting: An Electronic Health Record Investigation

Primary Care Diagnosis and Treatment of Attention-Deficit/Hyperactivity Disorder in School-Age Children: Trends and Disparities During the COVID-19 Pandemic

Rate of Pediatrician Recommendations for Behavioral Treatment for Preschoolers With Attention-Deficit/Hyperactivity Disorder Diagnosis or Related Symptoms

We aim to optimize operational aspects and work closely with clinicians to refine clinical workflows, improve care processes, and enhance overall service delivery through focusing on improving healthcare services at local clinics (quality improvement) and investigating best practices for integrating evidence-based practices for pediatric care (implementation science). While some of our work in this area is operational in nature and may not be classified as traditional research, we are dedicated to driving meaningful improvements in the care provided to children with developmental, behavioral, and mental health conditions.

We also integrate qualitative research methods, such as interviews, focus groups, and the Delphi process, to gain in-depth insights from key stakeholders, such as patients, families, healthcare providers, and other community members. By capturing their perspectives, experiences, and challenges in managing these conditions, we identify opportunities for care improvement. This qualitative data not only enriches our quality improvement efforts but also informs our AI and health services research, ensuring that our solutions are rooted in real-world clinical contexts and address the unique needs of those involved in the care process.