Background: With aging, the cognitive function of the prefrontal cortex (PFC) declined, postural control weakened, and fall risk increased. As a mind–body exercise, regular Tai Chi practice could improve postural control and effectively prevent falls; however, underlying brain mechanisms remained unclear, which were shed light on by analyzing the effect of Tai Chi on the PFC in older adults by means of functional near-infrared spectroscopy (fNIRS).Methods: 36 healthy older adults without Tai Chi experience were divided randomly into Tai Chi group and Control group. The experiment was conducted four times per week for 16 weeks; 27 participants remained and completed the experiment. Negotiating obstacle task (NOT) and negotiating obstacle with cognitive task (NOCT) were performed pre- and post-intervention, and Brodmann area 10 (BA10) was detected using fNIRS for hemodynamic response. A three-dimensional motion capture system measured walking speed.Results: After intervention in the Tai Chi group under NOCT, the HbO₂ concentration change value (ΔHbO₂) in BA10 was significantly greater (right BA10: p = 0.002, left BA10: p = 0.001), walking speed was significantly faster (p = 0.040), and dual-task cost was significantly lower than pre-intervention (p = 0.047). ΔHbO₂ in BA10 under NOCT was negatively correlated with dual-task cost (right BA10: r = −0.443, p = 0.021, left BA10: r = −0.448, p = 0.019). There were strong negative correlations between ΔHbO₂ and ΔHbR under NOCT either pre-intervention (left PFC r = −0.841, p < 0.001; right PFC r = −0.795, p < 0.001) or post-intervention (left PFC r = −0.842, p < 0.001; right PFC r = −0.744, p < 0.001).Conclusion: Tai Chi practice might increase the cognitive resources in older adults through the PFC bilateral activation to prioritize gait performance during negotiating obstacles under a dual-task condition.

Introduction: Chronic Ankle Instability (CAI) is a musculoskeletal condition that evolves from acute ankle sprains, and its underlying mechanisms have yet to reach a consensus. Mounting evidence suggests that neuroplastic changes in the brain following ankle injuries play a pivotal role in the development of CAI. Balance deficits are a significant risk factor associated with CAI, yet there is a scarcity of evidence regarding the sensorimotor cortical plasticity related to balance control in affected individuals. This study aims to evaluate the differences in cortical activity and balance abilities between patients with CAI and uninjured individuals during a single-leg stance, as well as the correlation between these factors, in order to elucidate the neurophysiological alterations in balance control among patients with CAI.Methods: The study enrolled 24 patients with CAI and 24 uninjured participants. During single-leg stance, cortical activity was measured using a functional near-infrared spectroscopy (fNIRS) system, which included assessments of the pre-motor cortex (PMC), supplementary motor area (SMA), primary motor cortex (M1), and primary somatosensory cortex (S1). Concurrently, balance parameters were tested utilizing a three-dimensional force platform.Results: Independent sample t-tests revealed that, compared with the uninjured individuals, the patients with CAI exhibited a significant increase in the changes of oxyhemoglobin concentration (ΔHbO) during single-leg stance within the left S1 at Channel 5 (t = 2.101, p = 0.041, Cohen’s d = 0.607), left M1 at Channel 6 (t = 2.363, p = 0.022, Cohen’s d = 0.682), right M1 at Channel 15 (t = 2.273, p = 0.029, Cohen’s d = 0.656), and right PMC/SMA at Channel 11 (t = 2.467, p = 0.018, Cohen’s d = 0.712). Additionally, the center of pressure root mean square (COP-RMS) in the mediolateral (ML) direction was significantly greater (t = 2.630, p = 0.012, Cohen’s d = 0.759) in the patients with CAI. Furthermore, a moderate positive correlation was found between ML direction COP-RMS and ΔHbO2 in the M1 (r = 0.436; p = 0.033) and PMC/SMA (r = 0.488, p = 0.016), as well as between anteroposterior (AP) direction COP-RMS and ΔHbO in the M1 (r = 0.483, p = 0.017).Conclusion: Patients with CAI demonstrate increased cortical activation in the bilateral M1, ipsilateral PMC/SMA, and contralateral S1. This suggests that patients with CAI may require additional brain resources to maintain balance during single-leg stance, representing a compensatory mechanism to uphold task performance amidst diminished lateral balance ability in the ankle joint.

Long short-term memory (LSTM) networks are widely used in biomechanical data analysis but have the significant limitations in interpretability and decision transparency. Combining graph neural networks (GNN) with gate recurrent units (GRU) may offer a better solution. This study proposes and validates a hybrid GNN-GRU model for predicting baseball pitching speed and enhancing its interpretability using layer-wise relevance propagation (LRP). C3D data from 53 baseball athletes were downloaded from a public dataset. Kinematic features of 9 joints and pitching speed during the pitching process were calculated using Visual3D, resulting in a total of 208 valid pitches. The feature data were input into both LSTM and GNN-GRU hybrid models, with hyperparameters tuned using particle swarm optimization. LRP was employed to obtain the contribution rate changes of kinematic features to the prediction results throughout the pitching cycle. The prediction accuracy of the models was evaluated using mean absolute error (MAE), mean squared error (MSE), and R-squared (R²). The results showed that there were the significant statistical differences in the MAE and R² metrics between the LSTM model and the GNN-GRU model in predicting pitching speed on the test set. The MAE (P = 0.000, Z = − 5.170, Cohen’s d = 1.514) and R² (P = 0.000, Z = − 2.981, Cohen’s d = 2.314) of the LSTM model were significantly lower than those of the GNN-GRU model. Compared to LSTM, the GNN-GRU model achieved better prediction accuracy but was potentially more susceptible to the influence of data variability. Moreover, the GNN-GRU-based model demonstrated the better interpretability and decision transparency.