Are 4D Motion Sensors Valid and Reliable for Studying Baseball Pitching?. Issue 6 (May 2023)

• Record Type: Journal Article
• Authors: Loushin, Stacy R.; Verhoeven, Martijn; Christoffer, Daniel J.; Camp, Christopher L.; Kaufman, Kenton R.
• Is Part Of: American journal of sports medicine. Volume 51:Issue 6(2023)
• Page Start: 1608
• Page End: 1614
• Publication Date: 2023-05
• Subjects: inertial measurement units -- motion capture; Sports medicine -- Periodicals; Sports injuries -- Periodicals; Orthopedic surgery -- Periodicals; 617.102705
• DOI: 10.1177/03635465231166423
• Languages: English
• ISSNs: 0363-5465

Abstract

Background: Baseball pitching injuries are on the rise. Inertial measurement units (IMUs) provide immediate feedback to players and coaches, allowing for collection outside of the traditional laboratory setting with real-world application. The 4D Motion system provides kinematics throughout the pitching motion and may be beneficial for individualized programs in the throwing athlete. A systematic analysis of these sensors has not been completed. Purpose: To evaluate the validity of the 4D Motion IMU system for analyzing the baseball pitching motion compared with marker-based motion capture, and evaluate the internal reliability and consistency of the device. Study Design: Controlled laboratory study. Methods: Ten high school pitchers participated in this study (10 male; 9 right-hand dominant; mean age, 16.6 ± 1.3 years; mean body mass index, 24.1 ± 3.9). Participants were simultaneously outfitted with six 4D Motion IMU sensors and retroreflective markers. The pitchers threw fastballs at maximum effort off a mound at the standard height and distance. A comparison was made between the IMUs and corresponding motion capture values for shoulder external rotation, elbow flexion, chest extension, pelvis and chest rotation velocity, and rotation acceleration. Results: Significant differences were found for 5 of 7 metrics analyzed. The IMU overreported most metrics, except for elbow flexion and pelvis rotation angular acceleration, where both positive and negative errors were observed. The root mean square error and percentage errors indicated smaller discrepancies for chest extension (4°± 5°) and pelvis (38 ± 19 deg/s) and chest (96 ± 42 deg/s) rotation velocity, with elbow flexion having the largest variance (21°± 9°). Conclusion: The values of the 4D Motion IMU system should not be considered equivalent when compared with marker-based motion capture studies. The system lacked internal consistency and reliability, with angular velocities being the most consistent. Caution should be used when using the metrics provided by an IMU-based system for individualized monitoring. Clinical Relevance: If found valid and reliable, IMUs could be used for longitudinal workload monitoring, individualized throwing and rehabilitation programs, and ultimately injury prevention. This study demonstrates that the data obtained from a 4D Motion system using Gen 3 sensors are not equivalent to the data obtained from a marker-based motion capture system.