Abstract
<title>Abstract</title> <p>Background This study evaluated the kinematic assessment capability of a low-complexity system, defined as one or two body-worn inertial measurement units (IMUs) using segment-level measures, during front squats, deadlifts, and back squats. Methods Forty-six adults completed 10 repetitions of each exercise under standardized submaximal loading and cadence-controlled conditions. IMUs were placed on the upper back and left thigh. Exercise effects were evaluated across 48 measures with false-discovery-rate control. Elastic-net multinomial regression with leave-one-participant-out validation compared upper-back-only, thigh-only, and combined configurations. Results Twelve kinematic measures differed among exercises. Seven were upper-back-derived: sagittal excursion; mean, 95th-percentile, and maximum recorded sagittal angular speed; sagittal velocity-profile ratio; and mean frontal- and transverse-plane angular speed. Three were thigh-derived: mean sagittal and transverse angular speed and sagittal velocity-profile ratio. Paired sagittal measures, normalized trunk-thigh excursion difference and trajectory correlation, were significant. No timing or repetition-consistency measure remained significant. Balanced accuracy was 65.2% for upper-back only, 42.8% for thigh only, and 63.0% for the combined configuration. Adding thigh-derived and paired measures did not improve performance over upper-back only in unseen participants. Conclusions Among the tested configurations, a single upper-back IMU provided the strongest exercise-specific assessment capability. The dual-IMU configuration added trunk-thigh information without improving exercise discrimination.</p>