Quantification of daily upper arm use after stroke using an IMU-based kinematic model

Upper limb (UL) impairment is one of the most common and disabling consequences of stroke, limiting independence and the performance of daily activities. Although rehabilitation aims to reduce impairments and enhance real-world activity performance, clinical motor assessments often fail to capture spontaneous UL use. This discrepancy between motor capacity observed in clinical settings and actual performance of daily activities underscores the need for objective, ecologically valid monitoring tools. Wearable inertial measurement units (IMUs) have emerged as feasible methodology for quantifying real-world UL use; however, the clinical interpretability and validity of kinematic metrics derived from measures of activity in persons with stroke (PwS) remain insufficiently established.

To evaluate the utility of three IMU-derived kinematic measures of UL use in persons with stroke (PwS) during routine daily activity and to examine their associations with clinical assessments of motor impairment.

Nineteen individuals in the subacute stage after stroke wore five IMU sensors during routine daytime activities in a rehabilitation hospital. Three kinematic measures - wrist path length, cumulative elbow angular displacement, and movement space volume - were extracted for each arm and combined to asymmetry indices. Clinical assessments included the Fugl-Meyer Assessment for the upper extremity (FMA-UE), Action Research Arm Test (ARAT), and grip strength. Asymmetry indices and their correlations with clinical measures were analyzed.

All IMU-derived measures were significantly reduced in the paretic arm compared with the non-paretic arm (p < 0.001). Asymmetry indices showed significant correlations with clinical impairment measures, with cumulative elbow angular displacement demonstrating the strongest association, particularly with ARAT scores.

IMU-based kinematic measures effectively capture reduced daily UL use in PwS and demonstrate associations with clinical impairment. These findings support the potential of wearable sensors to quantify daily arm use as an ecologically valid indicator of activity performance and recovery.