Acquired Cognitive Communication Disorders

5000V - Toward Personalized Treatment Recommendations: The Promise of Data-Driven Approaches in Speech Language Pathology

ON-DEMAND

Format: Technical Clinical (Virtual)

Audience: Students, Researchers

Keywords: acquired brain injury, machine learning, rehabilitation, predictive, mobile health data

Presenting Author(s)

Claire Cordella

Boston University

View Slides View Handouts

Author Only (Not Presenting)(s)

Hantian Liu

Boston University
Prakash Ishwar

Boston University
Margrit Betke

Boston University
Swathi Kiran

Boston University

This proposal aims to introduce an audience of clinicians and clinician-scientists to the fundamentals of data-driven modeling and machine learning and discuss their applicability to the field of communication sciences and disorders (CSD). We will present results drawn from analyses of a large dataset of patients with acquired brain injury who utilized a commercially available teletherapy application and discuss important patient- and treatment-level factors that show promise in predicting treatment-induced performance gains. This presentation will inform emerging discussions in CSD on the variability of individual treatment response profiles, the importance of personalizing treatment recommendations, and the role of data-driven modeling to help with these clinical challenges.

Learning Objectives:

Distinguish between explanatory and predictive modeling
Define machine learning
Discuss at least two clinical applications of machine learning and other data-driven approaches