Home → Research → Clinical informatics

AI and clinical informatics research

My work examines how clinical information can help psychiatrists assess patients, interpret evidence, and follow treatment over time.

I am an Assistant Professor of Clinical Psychiatry at Columbia University Irving Medical Center and Medical Director of Integrative Psych. My contributions span psychiatric epidemiology, clinical interpretation, and collaboration on software for mental health care.

MIND: narrative summaries of multimodal data

I am a coauthor of MIND: Empowering Mental Health Clinicians with Multimodal Data Insights through a Narrative Dashboard. The project explores how clinicians interpret information drawn from clinical notes and patient-generated data.

Research question
How can a narrative dashboard help clinicians interpret multimodal mental health information?
Study design
Co-design sessions with five clinicians, followed by a user study with 16 clinicians comparing the dashboard with baseline methods.
Reported outcome
Clinicians rated the dashboard more favorably for identifying relevant information and supporting decision-making.
Interpretation
The user study measures clinicians' perceptions of the interface. It does not establish improved patient outcomes or the performance of a separate practice deployment.
Source
Public paper and author record · Associated publication DOI

Clinical tools and interventions in development

Clinical decision support and CEBA-ADHD

This research direction examines how structured clinical knowledge can support ADHD assessment and organize information for clinicians. Evaluation priorities include diagnostic agreement, information accuracy, workload, and how clinicians respond to errors. A clinician remains responsible for assessment and treatment decisions.

PAWS

PAWS is an investigational digital intervention research project for young people with cannabis use disorder, developed with Xuhai “Orson” Xu, PhD. Its proposed functions require development and evaluation. Read the project overview.

Portable mental health records

A further research direction is a structured representation of diagnosis, treatment history, and outcomes that can support continuity of care and research. Development requires attention to the accuracy, provenance, and interpretation of information shared between systems.

Evaluation priorities

Studies of these tools need to specify the dataset, comparator, outcome, and setting. Interface usability, extraction accuracy, clinical decision-making, and patient outcomes are different endpoints and should be reported separately. Testing against designed crisis scenarios can identify failures; clinical safety also requires evaluation in the intended population and setting.

The MIND paper above documents a specific user study. The development projects described here should be understood as research directions whose clinical benefits remain to be evaluated. Research support is described separately.

Clinical research informing this work

My epidemiology research examines psychiatric prescribing and adolescent substance use. Those studies help identify clinical questions for informatics research.

Research contact

For collaboration on clinical informatics or digital intervention research, send a research inquiry.

Research program · Publications · Sultan Lab

Page updated: September 9, 2026