Capstone Project

Location: Washington University in Saint Louis

Areas of Focus: Evaluation and Research

Content Expert Mentor: Ganesh Babulal, OTD, PhD & Mario Millsap, OTD, OTR/L

Dates: May 2026 to August 2026

Purpose

The goal of this project is to develop skills in AI-focused research and identify implications of AI for occupational

therapy practice and research , ultimately improving the overall clinical experience in OT practice.

Summary of Capstone

This capstone was guided by literature review, mentor guidance, stakeholders interview and interview analysis, and data entry and data analysis.

Objectives

Objective 1: Enhance skills in AI-relevant research methods

1a. Guided readings to learn research approaches relevant to the team’s ongoing projects

1b. Conduct a preliminary analysis of data

1c. Describe implications for health research and practice.

Objective 2: Evaluate the ethical, practical, and user-centered considerations of implementing AI tools in OT practice.

2a. Review ethical guidelines and regulatory considerations for AI in healthcare regarding diversity, equity, and equality.

2b. Observe or assist in feedback, participatory design, or clinician interviews

2c. Identify socioeconomic barriers and facilitators to AI adoption and create recommendations.

2d. Shadow or interview OT clinicians/researchers/subjects (if have one) to understand current assessment challenges and opportunities for AI use.

Objective 3: Identify opportunities to expand use of AI in occupational therapy

3a. Observe or assist in feedback, participatory design, or clinician interviews

3b. Shadow or interview OT clinicians/researchers/subjects (if have one) to understand current assessment challenges and opportunities for AI use.

3c. Synthesize literature describing current use of AI in OT practice.

Research and Data

Washington University Drives Lab

Shadowed researchers conducting assessments with subjects.
Learned about assessment grading in vision tests, MOCA, and other dementia-related assessments.
Entered 40+ data points into REDCap for the longitudinal study, which helped me better understand the complexity of assessing driving ability.
Through conversations with lab members, I learned the potential for machine learning to organize messy data into analyzable datasets, which could ultimately solve more health related issues.
Presented capstone progress in weekly meetings and gathered feedback on the OT AI Literacy Toolbox.
Participated in a biweekly journal club to analyze different research papers. This experience helped me critically frame questions when reviewing resources and guided the development of AI-related educational materials for teaching OTPs to evaluate studies by critically analyzing recruitment, assessment choice, data analysis, and other key aspects of individual research designs. I also learned that results from poor study design can cause harm, and AI systems trained on such results may lead to ineffective treatments or an underestimation of patient needs.
Took notes and photographed during the CareSTL epilepsy focus group. This experience showed me how AI scribing can help in resource-limited settings. For example, note-taking throughout the day could be supported by AI scribing with consent, and the notes could then be organized by generative AI without PHI input. Under-resourced communities can benefit from technologies that support practice and strengthen community impact.