Classification of proper countermovement jumps using a sacral-mounted inertial measurement unit

Abstract The countermovement jump is an important neuromuscular assessment that drives sports performance and clinical decision-making. However, poorly executed countermovement jump techniques yield inflated or unreliable metrics when measured by force plates or inertial measurement units. This study aimed to develop an inertial measurement unit-based algorithm to classify countermovement jumps as proper or improper by identifying common technique errors. Ten participants performed three proper countermovement jumps and four intentionally improper jumps on force plates while wearing a sacral-mounted inertial measurement unit. The four improper jump variations included one each of the following: (1) a squat jump, (2) a jump with knees tucked during flight, (3) a jump with feet kicked out during flight, and (4) a deep squat landing. For the analysis, the jumps involving knees tucked and feet kicked out were grouped together under the category “flight movement.” The algorithm was developed to detect the improper jumps based on sacral inertial sensor acceleration and orientation. The inertial sensor algorithm detected 10/10 squat jumps, 18/20 jumps with flight movement, and 9/10 deep squat landings, while the force plates detected 4/10 squat jumps and were unable to identify other improper techniques. Binary proper/improper inertial measurement unit classifications achieved high sensitivity/recall (0.90), specificity (0.93), precision (0.90), and accuracy (0.91). These findings demonstrate that inertial sensors offer superior technique classification capabilities compared to force plates, and may enhance or replace them in clinical, athletic, and research settings to support more consistent countermovement jump assessments.