COMPARING A NOVEL SMARTPHONE APPLICATION TO THE KINECT V2 FOR ASSESSING ANTERIOR CRUCIATE LIGAMENT INJURY RISK WITH THE DROP VERTICAL JUMP

Screening for athletes at high risk for ACL tears who may optimally benefit from proven risk reduction training programs1 remains a challenge. Recent advancements have demonstrated that drop vertical jump (DVJ) parameters measured by a portable motion capture device (Kinect v2; Microsoft) can be a reliable risk assessment modality in collegiate varsity athletes.2 Development of a similar risk assessment modality utilizing a smartphone application would improve the simplicity, cost, and accessibility of this technology, allowing for widespread screening of high-level athletes. The purpose of the present study was to compare the DVJ tracking performance of a novel smartphone application to the previously validated3-5 Kinect v2 device in the context of ACL injury risk assessment. Two hundred fifty-two collegiate varsity athletes each performed three drop vertical jumps that were recorded simultaneously with an Android smartphone (Pixel 6; Google) and the Kinect v2. The smartphone application utilized software developed using the Google Mediapipe framework version 0.7. The Kinect v2 was paired to a computer with software developed using an open-source software development kit (Kinect v2; Microsoft). Initial coronal (IC), peak coronal (PC), and peak sagittal (PS) angles were identified by both systems. The intraclass correlation coefficient (ICC) test (two-way mixed-effects model, consistency, single measures) was used to assess the level of agreement between the smartphone application and the Kinect v2 for each parameter. Athletes were classified as high-risk or low-risk based on published cutoff values applicable to Kinect v2 measurements.2 Receiver operating characteristic (ROC) analysis was used to assess the accuracy of the smartphone application in identifying these high-risk athletes. Statistical significance was set at P < 0.05. ICC values for IC, PC, and PS angles captured between the smartphone application and the Kinect v2 were 0.696 [95% confidence interval (CI): 0.627 - 0.755, P The novel smartphone application demonstrates good to excellent agreement with the previously validated Kinect v2 device on specific drop vertical jump parameters that are amenable to ACL injury risk assessment. Further, the smartphone application identifies athletes classified as high-risk by the Kinect v2 system with excellent accuracy. The accessibility and simplicity of a smartphone application can enable widespread ACL injury risk screening. Further studies should examine the efficacy of the novel smartphone application in identifying athletes who subsequently suffer ACL tears.