Abstract
Existing clinical assessments for upper-limb motor rehabilitation poststroke pose limitations as end points for efficient clinical trials. This study aimed to develop a digital metric for assessing motor recovery using accelerometer data collected in naturalistic environments. We constructed the digital arm performance scale (DAPS) by analyzing ∼23,000 hours of data from 215 participants, including healthy individuals and subacute and chronic stroke survivors. We decomposed continuous upper-limb accelerometer data into lower-level movement segments, from which key features were extracted and aggregated using a linear mixed-effects model to produce an interpretable digital biomarker. DAPS demonstrated excellent reliability, sensitivity, concurrent validity, known-groups validity, discriminant validity, and responsiveness. Power analysis indicated that DAPS could reduce the required sample size for clinical trials with upper-limb motor recovery end points by more than 60% compared with traditional assessments. These findings highlight the potential of DAPS as a low-burden, scalable assessment tool for upper-limb motor recovery, with potential applications in both clinical trials and practice.