P 61 Robotic assessments of upper limb sensorimotor impairments in neurological disorders

Abstract

Background: Upper limb sensorimotor impairments are common in neurological disorders. In order to personalize and evaluate interventions aiming to improve upper limb function, it is essential to accurately assess the presence and extent of these impairments. Clinical assessments rely on experience-based observations, subjective evaluation, and ordinal scales, thus challenging a sensitive capture of intervention-induced changes in clinical trials. Novel robotic technologies promise sensitive and objective endpoints describing upper limb kinematic and kinetics during behavioural tasks. Over the last decades, we successfully developed, validated, and clinically applied such technologies in neurological disorders. The aim is to report on the feasibility, clinimetric properties, and clinical applicability of two complementary robotic assessments quantifying upper limb sensorimotor impairments. Methods: First, we relied on a haptic end-effector to record upper limb movement patterns and hand grip forces during a goal-directed pick-and-place task (Kanzler et al. 2020). Second, we custom-designed a robotic device and behavioural tasks providing precise stimuli to the index finger metacarpophalangeal joint, thereby allowing to study finger proprioception as well as motor and sensorimotor impairments (Zbytniewska et al. 2021). The recorded sensor data is analyzed with signal processing pipelines to generate objective digital health metrics describing movement quality (e.g., efficiency), grip force control (e.g., smoothness), and the response to well-controlled displacements (e.g., position matching error). Results: Feasibility of the assessments was established in persons with stroke, multiple sclerosis, or ataxia and matched able-bodied controls, revealing that the approaches are rapidly and easily applicable in a wide range of patients. For each platform, a core set of statistically optimal metrics was selected based on their test-retest reliability, measurement error, and discriminative power. Concurrent validity indicated low-moderate correlations with clinical scales, highlighting that the metrics provide novel, complementary information. By applying the metrics, we were able to reveal the presence of sensorimotor impairments and subtle longitudinal changes in persons that did not show abnormalities in clinical assessments. Conclusions: Robotic assessments are valuable tools to accurately and sensitively assess upper limb sensory, motor, and sensorimotor impairments in neurological disorders, thus opening novel avenues for evaluating clinical trials. References Kanzler CM, et al. A data-driven framework for selecting and validating digital health metrics: use-case in neurological sensorimotor impairments. npj Digit. Med. 2020;3(1):80. Zbytniewska M, et al. Reliable and valid robot-assisted assessments of hand proprioceptive, motor and sensorimotor impairments after stroke. J. Neuroeng. Rehabil. 2021;18(1):115.