.
Sports videos serve as critical resources for athletes and coaches to observe and replicate expert performance, yet current viewing practices lack quantitative tools to assess visual attention alignment with key body movements. To address this, we propose GazeSport, a visual analytics framework that integrates high-precision eye tracking with human pose estimation. GazeSport synchronizes gaze fixations and athlete key points to generate dynamic areas of interest (AOIs) using a region growing algorithm based on pose landmarks, adapting to motion in real time. It visualizes gaze behavior through spatiotemporal representations, allowing users to examine attention on relevant body parts and detect distractions toward nonessential elements such as referees and equipment. Viewing efficiency is evaluated using gaze entropy and Bayesian uncertainty. Designed for sports involving intensive body movement and suitable for individual or multi-athlete scenarios, GazeSport has been validated through a user study involving 18 volunteers with diverse professional backgrounds viewing the sports video from the SportsSloMo Dataset. Results reveal clear differences in attention efficiency among participant groups, demonstrating the effectiveness and generalizability of the proposed framework.
Resistance training is the cornerstone of sarcopenia management. However, whether combining it with amino acid-based supplementation, including essential amino acids, branched-chain amino acids, leucine, and derivatives such as HMB, provides additional benefits remains unclear. We systematically searched PubMed, Embase, Web of Science, and the Cochrane Central Register of Controlled Trials until February 2025. Meta- and subgroup analyses were performed using a random-effects model. The Cochrane ROB2 tool and GRADE framework were used to assess the risk of bias and evidence certainty. We included nine randomized controlled trials involving 496 participants with sarcopenia. Evidence rated as low to very low certainty indicated that, compared with resistance training alone, combined amino acid supplementation and resistance training significantly improved handgrip strength (SMD = 0.69, 95
Growing evidence implicates gut microbiome dysbiosis and its metabolites, particularly short-chain fatty acids (SCFAs), in the pathophysiology of autism spectrum disorder (ASD). However, individual study findings are inconsistent. Objective This systematic review and meta-analysis synthesized evidence comparing SCFA profiles between individuals with ASD and neurotypical controls. Methods We comprehensively searched PubMed, EMBASE, Cochrane Library, Web of Science, and CNKI databases from inception to December 1, 2024. Observational studies reporting quantitative SCFA measurements were included. Standardized mean differences (SMDs) with 95% confidence intervals (CIs) were calculated using random-effects models. Results Sixteen studies (473 ASD cases, 514 controls) were included. Meta-analysis revealed substantial heterogeneity across studies for most SCFAs. Despite this, levels of valeric acid (SMD = 0.76, 95% CI: 0.23 to 1.29) and hexanoic acid (SMD = 0.60, 95% CI: 0.05 to 1.16) were significantly elevated in individuals with ASD. Isobutyric acid also showed a positive association (SMD = 0.21, 95% CI: 0.03 to 0.46) with lower heterogeneity. No significant overall differences were found for acetic, propionic, or butyric acids, but subgroup analyses indicated crucial variations based on sample source (e.g., fecal, plasma, urine). Conclusion This meta-analysis provides evidence for altered SCFA profiles in ASD, specifically elevated valeric and hexanoic acids and a consistent signal for isobutyric acid, suggesting gut microbial dysbiosis involving distinct metabolic pathways. The significant heterogeneity and sample-source-dependent effects highlight the complexity of the gut-brain axis in ASD and underscore the need for future research with standardized protocols and longitudinal designs to clarify the role of SCFAs.
ABSTRACT As artificial intelligence (AI) and digital technologies continue to advance, their integration into higher music education reshapes traditional pedagogical approaches, particularly in sight‐singing and aural training, core components of musicianship. This systematic review examines empirical studies that explore the application, effectiveness and pedagogical integration of AI and digital tools to enhance these skills in higher education contexts. Following PRISMA guidelines, the review synthesises findings from diverse methodological approaches, ranging from quasi‐experimental designs to case studies. It highlights both AI‐powered platforms, such as intelligent tutoring systems and adaptive learning environments, and non‐AI digital tools, such as EarMaster and MIDI‐based systems. Results indicate that interactive and adaptive technologies significantly enhance pitch perception, rhythmic accuracy and melodic dictation when integrated within constructivist and blended learning models. However, critical limitations persist, including accessibility disparities, insufficient teacher training and the limited transparency of AI algorithms. The review identifies key gaps in longitudinal research, expressive skill development and cross‐cultural applicability. Findings aim to inform future curriculum design, policy development and innovation in AI‐assisted music pedagogy, supporting more inclusive, adaptive and pedagogically grounded approaches to aural and sight‐singing instruction.
BACKGROUND:Artificial intelligence (AI) and wearable sensors are increasingly reshaping sports injury risk prediction by enabling continuous, individualized, and data-driven assessment. DISCUSSION:This review summarizes recent advances in wearable technologies - including inertial measurement units (IMUs), electromyography (EMG), physiological monitors, and flexible electronics - and their use in capturing biomechanical, physiological, and psycho-physiological indicators relevant to injury risk. Deep learning (DL) models, particularly those capable of temporal and multimodal fusion, have shown promise in predicting injuries such as anterior cruciate ligament (ACL) tears, muscle fatigue, and stress fractures. Through representative applications and mechanism-informed analysis, we highlight the growing role of interpretable AI, real-time feedback, and behavior-integrated predictive systems. CONCLUSIONS:Key challenges remain, including data heterogeneity, limited generalizability, compliance issues, and privacy concerns. Future directions point toward personalized modeling, explainable systems, federated learning, and digital twin frameworks. Together, these advancements mark a shift toward proactive, intelligent injury prevention across athletic and clinical settings.
