Inertial measurement units (IMUs) offer a promising tool for assessing and monitoring performance in acrobatic gymnastics, potentially providing coaches with valuable biomechanical insights. IMU and motion capture data were collected synchronously in laboratory conditions while acrobatic gymnasts performed two acrobatic skills: a back tucked somersault (BTS) and a back straight somersault (BSS). Resultant linear acceleration and angular velocity (mediolateral axis) were calculated using data from an optical motion capture system (MOCAP) and using an IMU placed in the sacrum of 19 national and international level acrobatic gymnasts, divided into 11 pairs. The entire signals were compared using statistical parametric mapping. Results showed that the IMU systematically underestimated both acceleration and angular velocity when compared to the reference system. For the base gymnast's acceleration, four (BTS) and six (BSS) clusters showed significant differences. For the top gymnast, both tasks differed approximately 100% of the time (p < .001). Regarding angular velocity, eight (BTS) and three (BSS) clusters showed significant differences, reinforcing the underestimation observed. Our findings suggest that IMU data is not sensitive to the kinematic phases (speeds, accelerations, decelerations) occurring during the top gymnast flight. These findings suggest that a single IMU may not provide reliable or sufficient information for performance assessment in acrobatic gymnastics training.
© Copyright 2025 International Journal of Sports Science & Coaching. SAGE Publications. Alle Rechte vorbehalten.