Prognosticating global functional outcome in the recurrent ischemic stroke using baseline clinical and pre-clinical...

Area of Science:

• Neurology
• Medical Informatics
• Rehabilitation Medicine

Background:

• Recurrent ischemic stroke (RIS) leads to significant functional decline.
• Predicting global functional outcome (GFO) is crucial for effective rehabilitation planning.
• Early identification of GFO in RIS patients improves post-discharge care.

Purpose of the Study:

• To assess baseline features for classifying improving GFO in RIS patients.
• To identify patients with poor GFO at three months post-discharge.
• To develop a predictive model for GFO in recurrent ischemic stroke.

Main Methods:

• 86 RIS patients were divided into training and testing sets.
• Clinical and pre-clinical data were analyzed using machine learning algorithms.
• Support Vector Machine (SVM) was employed for classification and prediction tasks.

Main Results:

• The SVM model demonstrated high sensitivity and specificity in classifying GFO improvement.
• SVM accurately identified patients with poor GFO at three months post-discharge.
• The model showed robust performance on both training and testing datasets.

Conclusions:

• Machine learning models can effectively classify GFO responses in RIS patients.
• Predictive models aid in identifying patients with poor GFO for targeted interventions.
• This approach supports clinical decision-making in stroke rehabilitation.