International Journal of Physical Education, Fitness and Sports

Artificial Intelligence in Sports Science: Evidence Maturity, Athlete Digital Twins, Multimodal Analytics, and a Translation Agenda for 2026–2035

Akila S — 2026-08-29

rtificial intelligence is increasingly used in sports science, sports medicine, athlete monitoring, physical education, biomechanics, coaching, and performance analysis. However, many studies report technical success without demonstrating whether an approach is transportable beyond the development setting, useful in professional practice, or adequately protective of athlete interests. This systematic scoping review examined peer-reviewed research published from January 2016 to May 1, 2026 on machine learning, deep learning, computer vision, wearable and multimodal analytics, large language models, generative artificial intelligence, and athlete digital twins. Six bibliographic databases were searched, with Google Scholar and citation tracking used as supplementary discovery methods. The search identified 1,132 records; 901 records underwent title and abstract screening, 287 reports were assessed for eligibility, and 32 publications were included. Primary empirical and technical evidence was examined separately from review, conceptual, perspective, and governance literature. Ten fully verified primary empirical studies enabled direct study-level comparison. The strongest evidence came from narrowly defined tasks with explicit reference standards, including motor-skill assessment, biomechanical estimation, markerless motion capture, and athlete tracking. Evidence for injury prediction, wearable systems, generative AI, and athlete digital twins was less mature. Recurring limitations included small or homogeneous samples, single-dataset validation, limited calibration, incomplete reporting of model behaviour, restricted reproducibility, and a lack of prospective or independent evaluation. None of the included publications demonstrated sustained decision impact or attributable improvement in athlete, educational, sports-medicine, or organisational outcomes. Progress will require temporal and external validation, calibrated uncertainty, transparent reporting, representative datasets, interoperable multimodal systems, human oversight, and safeguards for athlete autonomy, privacy, fairness, and contestability.

Utilizing Active Video Games for Accurate Energy Expenditure Assessment in Physically Active Adults

Jorge A Aburto-Corona — 2026-07-15

This study validated the energy expenditure (EE) reported by the active video game (AVG) Fitness Boxing 2: Rhythm & Exercise by comparing it with indirect calorimetry (MC) in 50 university students (21.2 ± 2.0 years). Participants attended the laboratory on two occasions. During the first session, height (165.6 ± 9.1 cm), body weight (68.5 ± 15.6 kg), and maximal oxygen consumption (42.2 ± 8.2 mL/kg/min) were assessed using a progressive treadmill test. In the second session, participants completed 15 minutes of free-mode exercise on the AVG while connected to the metabolic cart, allowing for simultaneous collection of EE from both the game and indirect calorimetry. Heart rate (162.0 ± 20.5 bpm), perceived exertion (11.5 ± 2.5), and enjoyment (73.3 ± 7.3) were also measured. The results showed significant gender differences in EE (p = 0.049). Specifically, EE estimates differed between devices in women (AVG: 110.3 ± 25.3 kcal vs. MC: 95.3 ± 25.3 kcal; p = 0.003). Although no significant differences were observed in men, the moderate correlation coefficient (r = 0.60) and the standard error of measurement (19.5 kcal), which exceeded the minimal detectable change (5.5 kcal), indicate substantial measurement error. Overall, these findings suggest that the AVG presents considerable inaccuracy when estimating EE and therefore should not be used in clinical or diagnostic contexts. However, it may still serve as a useful tool for promoting exercise participation and adherence.

Effects of Project-Based and Deep Learning Pedagogy on Dribbling Skills in Physical Education Students: A Quasi-Experimental Study

Ramadhan Arifin — 2026-08-15

This study examined the effectiveness of integrating Project-Based Learning (PjBL) with a deep learning approach to improve university students’ dribbling skills. A quantitative quasi-experimental method with a nonequivalent control group pretest–posttest design was employed. The participants were 70 students enrolled in the Physical Education Study Program at Universitas Lambung Mangkurat, consisting of 35 students in the experimental group and 35 in the control group. The experimental group received PjBL integrated with mindful, meaningful, reflective, collaborative, and joyful learning principles, whereas the control group received conventional instruction. Dribbling performance was assessed using an instrument adapted from the Modified Illinois Change-of-Direction Test with Ball, measuring change-of-direction speed, ball control, foot–ball contact, visual attention, and the ability to avoid contact with cones. The experimental group’s mean score increased from 69.09 at pretest to 79.69 at posttest, while the control group’s mean score increased from 69.09 to 71.31. Analysis of covariance, with posttest score as the dependent variable and pretest score as the covariate, revealed a significant group effect after controlling for initial ability, F(1, 67) = 287.29, p < 0.001, partial η² = 0.811. The adjusted posttest mean was significantly higher in the experimental group (M = 79.69, SE = 0.35) than in the control group (M = 71.31, SE = 0.35), with an adjusted mean difference of 8.37 points, 95% CI [7.39, 9.36]. These findings indicate that integrating PjBL with a deep learning approach produced a substantial improvement in university students’ dribbling skills.