Measuring Vertical Jump Height With Artificial Intelligence Through a Cell Phone: A Validity and Reliability Report.

This study estimated the reliability and validity of an artificial intelligence (AI)-driven model in the My Jump 2 (My Jump Lab) for estimating vertical jump height compared with the Force Platform (FP). The cross-sectional study involved 88 athletes (33 female and 55 male athletes), performing a total of 264 countermovement jumps with hands on hips. "Jump heights were simultaneously measured using the FP and the My Jump2app." The FP estimated jump heights using the impulse-momentum method, whereas My Jump 2 used the flight-time method, with the latter using an AI feature for automated detection of jump take-off and landing. Results indicated high reliability for the AI model (intraclass correlation coefficient [ICC1.3] = 0.980, coefficient of variation [CV] = 4.12) and FP (ICC1.3 = 0.990, CV = 2.92). Validity assessment showed strong agreement between the AI model and FP (ICC2,k = 0.973). This was also supported by the Bland-Altman analysis, and the ordinary least products regression revealed no significant systematic or proportional bias. The AI-driven model in My Jump 2 is highly reliable and valid for estimating jump height. Strength and conditioning professionals may use the AI-based mobile app for accurate jump height measurements, offering a practical and efficient alternative to traditional methods. [ABSTRACT FROM AUTHOR]