A New Robotic System Evaluates Post-Stroke Mobility

Stroke stands as a major contributor to long-term disability on a global scale. Annually, over 15 million individuals worldwide suffer strokes, with three-quarters of survivors encountering arm and hand impairments, weakness, or paralysis.

Post-stroke, many rely predominantly on their stronger arm for everyday activities, even when their weaker arm could regain function. This tendency, termed “arm nonuse” or “learned nonuse,” hinders recovery and may lead to injury.

Measuring the use of the weaker arm outside clinical settings poses a challenge due to the observer’s paradox—requiring covert assessment for natural behavior. Researchers at USC have introduced a unique robotic system designed to track arm usage in stroke survivors precisely.

Their pioneering method, detailed in Science Robotics’ November 15 publication, employs a robotic arm for 3D spatial tracking and machine learning algorithms to gauge “arm nonuse.” A socially assistive robot (SAR) guides and motivates participants during the evaluation.

A new robot assesses post-stroke mobility: a stroke survivor’s progress

Nathan Dennler, the lead author and a doctoral student in computer science, highlighted the goal: to evaluate how a stroke survivor’s progress in physical therapy translates to real-life scenarios.

Researchers from USC’s Thomas Lord Department of Computer Science and the Division of Biokinesiology and Physical Therapy collaborated on this study. Maja Matari?, a study co-author and Chan Soon-Shiong Chair and Distinguished Professor of Computer Science, Neuroscience, and Pediatrics, noted the integration of quantitative user-performance data collected by a robot arm.

This data collection, paired with a socially assistive robot, aims to motivate users to exhibit representative performance. The combination is seen as a more precise and engaging approach to assessing stroke patients.

Other contributors include Stefanos Nikolaidis, an assistant professor of computer science; Amelia Cain, an assistant professor of clinical physical therapy; Carolee J. Winstein, a professor emeritus and an adjunct professor in the Neuroscience Graduate Program, alongside computer science students Erica De Guzmann and Claudia Chiu.

A new robot assesses post-stroke mobility: Mirroring everyday use

The research team enlisted 14 participants who were initially right-hand dominant before experiencing a stroke. These individuals positioned their hands on a 3D-printed box with touch sensors, which served as the device’s home position.

A socially assistive robot (SAR) explained the system’s mechanics and offered encouraging remarks. Simultaneously, a robot arm maneuvered a button to various target locations in front of the participant, totaling 100 locations. The “reaching trial” commenced when the button was illuminated, signaled by the SAR instructing the participant to act.

During the initial phase, participants were encouraged to reach for the button using their preferred hand, replicating their everyday practice. In the next stage, they used only the affected arm, mimicking actions often observed in therapy or clinical settings.

Employing machine learning, the team assessed three parameters. Arm use probability, time to reach, and successful reach—to calculate an arm nonuse metric. A significant contrast in performance between the phases would imply the underutilization of the affected arm.

“The participants face a time constraint to reach the button, so despite being aware of the assessment, they must react swiftly,” Dennler explained. “This allows us to gauge an immediate response to the light signal—deciding in an instant which hand to use.”

Safe and easy to use

The researchers noted variations in hand choice and reaching time among chronic stroke survivors. Their method proved consistently reliable, offering a user-friendly experience with positive safety ratings from participants.

Crucially, they identified distinct patterns in arm use, potentially aiding healthcare professionals in monitoring stroke recovery more accurately.

“One participant showed reduced right-arm usage in upper areas on that side but higher usage for lower regions,” Dennler explained. “Another relied more on their less-affected side for upper points closer to the midline.”

This technology offers comprehensive insights for rehabilitation therapists, helping tailor interventions based on a patient’s specific needs and strengths.

Read the original article on sciencedaily.

Reference: Nathaniel Dennler, Amelia Cain, Erica De Guzmann, Claudia Chiu, Carolee J. Winstein, Stefanos Nikolaidis, Maja J. Matarić. A metric for characterizing the arm nonuse workspace in poststroke individuals using a robot arm. Science Robotics, 2023; 8 (84) DOI: 10.1126/scirobotics.adf7723.

Read more: 2023 Industrial Design Trends to Watch