Publication

Abstract : Affordable wearable inertial measurement unit (IMU) gloves can record wrist motion outside the laboratory. However, consumer designs expose no forearm landmarks (humeral epicondyles, radial and ulnar styloid processes). The wrist axes are therefore set from rig rest-pose geometry, a geometric proxy that need not align with the true anatomical rotation axis. We tested whether a commercial glove system (Rokoko Smartgloves) can measure within-limb wrist flexionextension (FE) and radial-ulnar deviation (RUD) coupling, benchmarked against published optical values. Thirty-one adults performed bilateral FE and RUD active range-of-motion (ROM) tasks. The glove’s default FE axis lay 35-39 degrees from the functionally calibrated axis (left hand 39 degrees, right hand 35 degrees; between-participant SD about 8 degrees). In this uncorrected frame the two axes appeared almost perfectly coupled (absolute r about 0.99). Functional axis calibration removed this kinematic crosstalk artifact. FE-task coupling fell to near zero (absolute r 0.01 to 0.10), whereas RUD-task coupling persisted (absolute r 0.41 to 0.61) in the dart-thrower direction on both hands. Off-axis leakage was present in every task and greater for RUD-dominant than FE-dominant movements, matching Li’s asymmetry. Peak ROM also matched optical norms. These results confirm that functional axis calibration is a prerequisite for using entry-level commercial IMU gloves in wrist biomechanics.