Motor-cognitive functions required for driving in post-stroke individuals identified via machine-learning analysis

Area of Science:

• Neuroscience
• Rehabilitation Medicine
• Artificial Intelligence

Background:

• Stroke survivors often face challenges with driving, necessitating accurate driving aptitude evaluation.
• Understanding the specific motor-cognitive functions crucial for driving is essential for post-stroke safety.
• Current methods for assessing driving aptitude in this population require refinement.

Purpose of the Study:

• To develop a machine learning algorithm for automatically selecting key motor-cognitive indices related to driving aptitude.
• To identify specific motor-cognitive functions critical for driving in individuals who have experienced a stroke.
• To provide evidence-based criteria for determining the need for on-road driving tests.

Main Methods:

• A machine learning algorithm with sparse regularization was applied to 65 motor-cognitive indices from 10 tests.
• The study involved 55 participants who had previously been hospitalized with stroke.
• The algorithm identified significant indices predictive of driving aptitude.

Main Results:

• The proposed method achieved high accuracy (AUC 0.946) in predicting driving aptitude.
• A specific set of motor-cognitive function indices strongly correlated with driving aptitude was identified.
• The analysis clarified the relationship between motor-cognitive function and driving ability.

Conclusions:

• The algorithm effectively identifies driving-related motor-cognitive functions from complex test data.
• This approach enables autonomous evaluation of driving aptitude in post-stroke individuals.
• The method has the potential to reduce screening tests, clinical workload, and enhance safety for stroke survivors.