The Potential Predictors of Motor Performance Outcomes after Rehabilitation for Patients with Stroke

Abstract

The identification of potential predictors for motor outcome after rehabilitation helps underscore the factors that may affect treatment outcomes and target individuals who benefit the most from the therapy. In this study, we addressed and utilized a classifier to identify the potential predictors for motor performance outcome for patients with stroke after rehabilitation. The potential predictors selected and used by different assessments in this study were age, sex, time since stroke, education, neurologic status, and the movement performance of the upper extremity. This study aimed to identify predictors of motor performance outcomes after rehabilitation for stroke patients. The PSO-SVM was chosen in this study to find the predictor of motor function for clients with stroke. The potential predictors for motor outcome after rehabilitation were motor ability assessment of the Fugl-Meyer Assessment (FMA) and the Functional Independence Measure (FIM). This study is to investigate the potential demographic and clinical characteristics of stroke that can serve to predict rehabilitation outcomes in motor performance.

Investigating the structure-function relationship of the corticomotor system early after stroke using machine learning

“…Support vector machine is a commonly used machine learning algorithm that solves binary classification problems by calculating a decision boundary. Support vector machine has previously been used to predict future motor outcomes and recovery in the upper limb after stroke using demographic, clinical, imaging, and neurophysiological variables ( Lee et al, 2013 , Rehme et al, 2015 , Guggisberg et al, 2017 ). Support vector machine can also be used for cross-sectional investigation of corticomotor structure–function relationships by using MRI metrics of corticomotor structure to classify a measure of corticomotor function such as MEP status.…”