L2M – Designing and validating an AI-based, user-centered tool to assess discourse skills in individuals with aphasia

Up to 40% of stroke survivors can experience aphasia, a language disorder that affects a person’s ability to speak, understand, read, or write. However, one of the most effective ways to accurately assess aphasia, discourse analysis, remains underutilized. Discourse analysis is usually time-consuming and performed manually, leading to varying results between clinicians. It can also be difficult to integrate into busy clinical settings, leaving many stroke survivors in need of this assessment method without access. Practical assessments like discourse analysis are important because they help clinicians choose the most appropriate therapy options and track patient recovery more efficiently.

The purpose of this project is to develop an automated discourse analysis tool powered by Artificial Intelligence that is designed to assess linguistic difficulties in individuals with aphasia. This tool will be used to transcribe aphasic speech, extract its linguistic features, and deliver results in minutes, not hours. Designed by clinicians for clinicians, this tool can be easily used in busy clinical settings, providing clinicians with quick, accurate, and reliable assessments that can support diagnosis, treatment planning, and outcomes. By using this tool, clinicians may save time and reduce variability in their assessments, allowing them to focus on patient care. During this internship, we will conduct the first phase of the project, which includes engaging with potential users like speech-language pathologists, to gain insights into how to design and further develop this tool for the next phases of the project.

In terms of benefits at the health systems level, this tool may reduce wait times, optimize the use of resources, and improve the overall quality of rehabilitation services for patients. By making discourse analysis fast, consistent, and accessible, this tool could support its wider clinical use for individuals with aphasia.

Faculty Supervisor:

Tanya Dash

Student:

Partner:

DMZ Ventures Inc

Discipline:

Life Sciences

Sector:

Professional, scientific and technical services

University:

University of Alberta

Program: