Artificial intelligence (AI), virtual reality (VR), and exergame technologies have been increasingly used in dance and movement activities. However, this literature remains dispersed across different areas, making it difficult to determine how the field has developed. This study mapped the research landscape and thematic development of AI-, VR-, and exergame-based dance and movement research on psychological outcomes using bibliometric analysis and latent Dirichlet allocation (LDA) topic modeling. A total of 252 records indexed in the Web of Science Core Collection from 2011 to 2025 were included. Five related thematic strands were identified: immersive dance interaction and technology-supported teaching; rehabilitation-oriented dance or rhythm training; school-based exergaming and psychophysiological assessment; behavioral program design and intervention implementation; and AI-based motion or emotion recognition. These strands indicate that the field has developed into a multi-layered research space shaped by technology functions, movement contexts, intervention formats, and psychological constructs, rather than a single dance-intervention or technology-application domain. At the same time, psychological outcomes were not represented with equal clarity across these strands. Participation-related and psychosocial constructs, including enjoyment, motivation, engagement, self-efficacy, social interaction, emotional expression, and quality of life, were more frequently represented, whereas mental-health-related outcomes such as anxiety, depression, stress, loneliness, and psychological well-being were less consistently connected to technology-supported dance or movement interventions. These findings clarify where evidence is concentrated, how major themes are organized, and where psychological outcome measurement requires clearer theoretical and methodological specification. Future studies should use comparative and longitudinal designs to examine whether VR/AI-based feedback-supported movement training offers added value over conventional dance or movement programs for psychological outcomes, participation, exercise experience, and longer-term behavior change.

Virtual reality (VR) has been increasingly adopted as a digital tool in rehabilitation for balance training, coordination improvement, and motor recovery, yet the literature remains dispersed across clinical rehabilitation, exercise-based interventions, and broader motor-related applications. This fragmentation makes it difficult to determine how the field has evolved and where research emphasis has shifted. This study mapped the research landscape and thematic evolution of VR for balance, coordination, and motor rehabilitation using bibliometric analysis and topic modeling. A total of 1258 articles indexed in the Web of Science Core Collection from 2011 to 2025 were analyzed. Only English language articles and reviews relevant to VR-based balance, coordination, or motor rehabilitation research were included, yielding a final dataset of 1258 publications. CiteSpace and VOSviewer were used to examine keyword co-occurrence, clustering patterns, and temporal trends, while Latent Dirichlet Allocation (LDA) was applied to identify latent themes and their temporal dynamics. The field has moved beyond early feasibility testing toward a more differentiated landscape shaped by distinct clinical targets, population groups, and training purposes. Seven recurring themes were identified, including vestibular rehabilitation and immersive training, post-stroke upper-limb rehabilitation, efficacy and adverse-effect assessment, balance and gait training interventions, evidence synthesis and review-based evaluation, elderly exercise and cognitive interventions, and skill-oriented virtual task training with recent expansion toward broader population groups and task-specific applications beyond traditional rehabilitation settings. VR research on balance, coordination, and motor rehabilitation has evolved into a more thematically differentiated field rather than remaining a single rehabilitation-oriented domain. By combining bibliometric mapping with topic modeling, this study clarifies where evidence is concentrated and which thematic directions are gaining visibility, providing a clearer basis for future evidence synthesis and more comparable intervention reporting.

Artificial intelligence (AI) has rapidly reshaped sports motion analysis through advances in wearable sensing, computer vision, and deep learning. However, existing reviews often focus on isolated techniques and lack a systematic evolutionary perspective. This study presents a 15-year bibliometric and evolutionary review of AI in sports motion analysis (2011–2025), integrating scientometric mapping with quantitative content analysis across modalities, methods, and applications. A comprehensive multi-source dataset of 2602 publications was analyzed using VOSviewer and CiteSpace to examine knowledge structures, collaboration patterns, co-citation networks, and emerging research fronts. In addition, each study was categorized by modality (wearable, visual, and multi-modal), AI method (traditional machine learning, deep learning, and transformer-based), and task type (activity recognition, performance analysis, rehabilitation, and others). The results reveal exponential growth and a clear three-stage evolution: a sensor-driven phase dominated by wearable devices and traditional machine learning (2011–2015), a deep learning expansion phase centered on vision-based modeling (2016–2019), and a recent deep learning consolidation phase characterized by emerging transformer-based methods, increasing multimodal integration, real-time monitoring, and application-oriented sports analytics (2020–2025). The field has shifted from signal-based recognition toward vision-centered performance evaluation, rehabilitation-related movement assessment, and injury-informed applications. This data-driven review provides an integrated evolutionary framework and future research roadmap to support the continued development of AI-driven analytics in sports science, athlete monitoring, and health-related human movement analysis.

Balance and postural control are important for safe movement in college students. This six-week randomized trial compared an HTC Vive Cosmos-based gamified virtual reality (VR) balance program with traditional balance and coordination exercises in 62 college students (VR: n = 32; control: n = 30). Both groups trained three times per week for 30 min per session. Outcomes included single-leg stance with eyes closed (SLST-EC; primary), Y-Balance Test composite score (YBT; secondary), and rotary stability (secondary), assessed before and after the intervention. Both programs were associated with improvements over time. The VR group showed greater improvement than the control group in SLST-EC, with a large between-group effect (Cohen’s d = 1.245), and this between-group difference remained significant after adjustment for baseline values. For YBT, both groups improved, but no clear overall between-group advantage was observed after accounting for baseline differences. Rotary stability also improved in both groups, without a significant between-group difference. Overall, gamified VR showed the clearest between-group advantage for static balance and may serve as a feasible supplement to balance training in college settings.

Objective assessment of technical execution in artistic gymnastics remains challenging because scoring relies on expert visual judgement of posture, movement control, alignment, symmetry, and action-specific execution quality. This study developed a markerless, video-based, and biomechanically interpretable machine learning framework for estimating gymnastics execution scores from pose-derived movement features. A total of 500 men’s artistic gymnastics floor-element video sequences were collected under controlled conditions and independently scored by three FIG-certified judges, with the mean score used as the ground-truth label. Pose trajectories were extracted using MediaPipe Pose, and judging-relevant landmarks were retained to represent key body segments. After hip-centred coordinate normalisation, biomechanical features were constructed from four groups: position magnitude, joint angle, temporal velocity, and left–right symmetry. As the core prediction method, this study proposes a Biomechanical Feature-Group Attention Support Vector Regression model (BFG-AttSVR). BFG-AttSVR introduces an action-specific feature-group attention mechanism to learn different group-level importance weights for different gymnastics actions. Within BFG-AttSVR, the attention weights are learned through a two-stage optimisation strategy, including data-driven initialisation based on cross-validated Ridge regression and gradient-free fine-tuning using an RBF-kernel SVR objective. Experimental results showed that BFG-AttSVR achieved an MAE of 0.0465 ± 0.0041, an RMSE of 0.0621 ± 0.0074, and an $R^2$ of 0.9710 ± 0.0131, outperforming the strongest SVR (RBF) baseline with an RMSE of 0.0743 ± 0.0071. The proposed framework provides a low-cost, objective, and interpretable approach for video-based gymnastics score estimation.

Balance and coordination are fundamental for children’s motor development and injury prevention, yet sedentary lifestyles have contributed to their d