Symposium

Towards Precision Rehabilitation for Multiple Sclerosis: Identifying Responsiveness to Virtual Reality Rehabilitation Interventions Using Machine Learning 0587

Corresponding Presenter(s)

• 

  Hannes Devos, PhD (he/him/his)

  Associate Professor
  University of Kansas Medical Center
  Overland Park, KS, United States

Presenter(s)

• 

  Maryam Sadeghi, MSc (she/her/hers)

  GRA_ PhD Candidate
  University of Kansas Medical Center
  Overland Park, Kansas, United States
• 

  Jacob J. Sosnoff, PhD

  Professor and Associate Dean for Research
  University of Kansas Medical Center
  Kansas City, Kansas, United States

Although physical rehabilitation and cognitive training are effective for individuals with multiple sclerosis (MS), the response to treatment is highly variable. By examining two recent randomized controlled trials (RCTs) that compare virtual reality (VR) interventions with traditional methods, we aim to explore how machine learning (ML) techniques, including artificial neural networks and decision tree models, can effectively classify MS patients as responders or non-responders. Such ML techniques to identify treatment response are urgently needed to enhance the tailor interventions to individual response patterns

Learning Objectives:

• Explain the variability in MS patient responses to VR-based physical and cognitive rehabilitation interventions.
• Recognize the role and efficacy of VR interventions in improving MS-related motor and cognitive functions.
• Explore the application of ML techniques, including artificial neural networks, K-nearest neighbor, and decision tree models, in predicting treatment outcomes and personalizing rehabilitation strategies.