Introduction: Running is one of the most popular sports in the world, but it also increases the risk of injury. The purpose of this study was to establish a modeling approach for IMU-based subdivided action pattern evaluation and to investigate the classification performance of different deep models for predicting running fatigue.Methods: Nineteen healthy male runners were recruited for this study, and the raw time series data were recorded during the pre-fatigue, mid-fatigue, and post-fatigue states during running to construct a running fatigue dataset based on multiple IMUs. In addition to the IMU time series data, each participant’s training level was monitored as an indicator of their level of physical fatigue.Results: The dataset was examined using single-layer LSTM (S_LSTM), CNN, dual-layer LSTM (D_LSTM), single-layer LSTM plus attention model (LSTM + Attention), CNN, and LSTM hybrid model (LSTM + CNN) to classify running fatigue and fatigue levels.Discussion: Based on this dataset, this study proposes a deep learning model with constant length interception of the raw IMU data as input. The use of deep learning models can achieve good classification results for runner fatigue recognition. Both CNN and LSTM can effectively complete the classification of fatigue IMU data, the attention mechanism can effectively improve the processing efficiency of LSTM on the raw IMU data, and the hybrid model of CNN and LSTM is superior to the independent model, which can better extract the features of raw IMU data for fatigue classification. This study will provide some reference for many future action pattern studies based on deep learning.

Background. There is a possible interaction between the underlying mechanisms of perceptual-cognitive tasks and motor control. For example, landing biomechanics changed during perceptual-cognitive tasks undertaken at different levels of fatigue of the athlete. Thus, this study explored the effect of perception-cognitive tasks interventions on male soccer players landing mechanisms at different levels of fatigue. Methods. Perceptual-cognitive tasks during games were simulated using classic multiple object tracking (MOT) paradigms, and 15 male soccer players completed MOT tasks under nonfatigue (NF), moderate fatigue (MF), and severe fatigue (SF). Landing-associated indicators were collected and calculated using a Vicon and force measuring platform. Results. Level of fatigue and MOT task significantly affected hip and knee flexion angles, hip and knee extension moments, and vertical ground reaction force. Specifically, hip and knee flexion angles were significantly higher in MOT than non-MOT tasks at all levels of fatigue. In NF state, hip and knee extension moments were significantly smaller during MOT than non-MOT tasks. In SF state, the hip extension moment was larger during MOT than non-MOT tasks. In both MF and SF states, vertical ground reaction force was significantly higher in MOT than non-MOT tasks. Conclusion. Although soccer players landed cautiously when not fatigued, they were significantly less able to do this and handle challenging perceptual-cognitive task movements when fatigued. Thus, landing performance is affected by perceptual-cognitive task interference in fatigue conditions.

Introduction: Tai Chi has been shown to exert beneficial effects on cognitive function in older adults; however, the mechanisms underlying its improvement of postural control and mobility remain unclear. This study aimed to investigate the effects of a 3-month Tai Chi training intervention on cognitive function and postural control in older adults, thereby providing a scientific basis for exercise recommendations for this population. Method: This study aimed to evaluate the effects of 8 forms of Tai Chi training over a period of 3 months on gait characteristics. A total of 34 older adults (25 females) completed the full study protocol. Walking kinematic data were collected using the MyoMotion 3D motion analysis system under single-task and three dual-task conditions (walking while counting, carrying water, and naming). Spatiotemporal parameters, including step length, cadence, and lower limb joint angles, were extracted through gait cycle analysis. The control group maintained their habitual daily activities. Repeated measures analysis of variance (ANOVA) was used to compare differences between groups and across time points (pre- and post-intervention), with the significance level set at 0.05. Result: Baseline characteristics were comparable between groups (p >0.05). Post-intervention, the experimental group demonstrated significant improvements in gait spatial parameters, characterized by increased normalized step and stride lengths and decreased step width compared to controls (p <0.05). Temporally, the experimental group exhibited an increased percentage of stance phase. Kinematically, the intervention significantly enhanced hip range of motion (p <0.01) and altered knee and ankle joint angles at specific gait events. Notably, a significant group-by-task interaction was observed for maximum hip extension (p <0.05), indicating that Tai Chi training differentially optimized hip mechanics across varying dual-task conditions. Additionally, the experimental group showed significantly improved overall cognitive function (p < 0.01). Conclusion: The Tai Chi intervention effectively improved gait stability, balance, and overall gait performance under dual-task conditions in older adults with MCI, thereby reducing the risk of falls. These findings support Tai Chi as a safe, cost-effective, and scalable non-pharmacological intervention for fall prevention in this population.