Journal of NeuroEngineering and Rehabilitation最新文献

Background: Patients undergoing anterior cruciate ligament reconstruction (ACLR) are at high risk of osteoarthritis or secondary injuries, with abnormal knee contact forces (KCFs) identified as a key factor in joint degeneration. Traditional KCF assessment relies on expensive lab systems while advances in computer vision and AI now enable low-cost alternatives. However, currently available methods oversimplify knee mechanics and neglect compensatory movements, highlighting the urgent need for intelligent, real-time monitoring tools for personalized rehabilitation. Therefore, the aim of this study was to develop and validate an integrated, non-invasive framework for accurate KCFs prediction in ACLR patients during daily activities. We hypothesized that combining enhanced musculoskeletal modeling with a deep learning architecture incorporating spatiotemporal attention would improve the prediction accuracy across multiple movement tasks.

Methods: This study simultaneously recorded three daily movements of 29 post-ACLR patients using both Vicon and OpenCap. Motion trajectories captured by Vicon were imported into OpenSim for musculoskeletal modeling and KCFs calculation. Dataset comprising OpenCap-derived kinematics and OpenSim-computed KCFs was used to train 3 learning models for the prediction of KCFs in ACLR patients across different movements.

Results: Among three models, CNN-BiGRU-Attention model demonstrated the best predictive performance across all three movement tasks (R2walking = 0.973 ± 0.003, R2running = 0.982 ± 0.004, R2descending stairs = 0.951 ± 0.007). CNN and self-attention mechanism collectively enhanced the model's ability to capture key features in ACLR patients' movement data, thereby improving KCF prediction accuracy. Furthermore, for the three daily activities, all models showed superior KCFs prediction performance in running and stair-descent tasks compared to walking.

Conclusion: The developed framework successfully achieved high-precision prediction of KCFs. This technological breakthrough not only provides a real-time quantitative tool for rehabilitation monitoring in patients with ACLR, but also facilitates a paradigm shift from static laboratory analysis to dynamic real-time monitoring, with broad application prospects in sports medicine, rehabilitation engineering.

Background: Late-onset Pompe disease (LOPD) presents with progressive muscle weakness, often leading to functional impairment that is challenging to monitor with conventional assessments. This study aims to develop and validate novel gait-based outcome measures for monitoring disease progression in individuals with LOPD.

Methods: Longitudinal study with genetically confirmed LOPD patients and age-and gait velocity-matched healthy controls that were assessed over a two-year period using the Ephion Mobility system, which integrates inertial sensors, plantar pressure insoles, and surface electromyography. All participants completed a free walking test (10-15 m at self-selected pace) and the 6-minute walk test (6MWT). Differences in gait features were identified using a three-stage feature selection framework that includes linear mixed-effects model, ElasticNet-regularized and bootstrap analysis. To explore intra-group variability within the LOPD cohort, we performed a clustering analysis. Based on the selected features and their weighted temporal changes, we developed a Pompe Mobility-Derived Progression Index (Pompe-MDPI) by training a Linear Discriminant Analysis (LDA) to discriminate between control and LOPD data. We calculated the Minimum Clinically Important Difference and compared its performance against the 6MWT distance.

Results: 24 LOPD and 39 healthy controls were included in the study. 46 gait features were found to significantly differentiate individuals with LOPD from controls (Holm-corrected p < 0.05), comprising 16 from trunk and pelvis joints, 18 from lower limb joints, 4 from force profiles, and 8 from EMG.Hierarchical clustering analysis revealed two distinct subgroups within the LOPD cohort, based on nine gait features. The computed Pompe-MDPI successfully discriminated between LOPD and healthy controls (AUC = 0.95), outperforming the 6MWT distance (AUC = 0.84). The Pompe-MDPI was also strongly associated with the 6MWT (p < 0.0001) and demonstrated significant change over time in the LOPD group (p = 0.02).

Conclusions: The Pompe Mobility-Derived Progression Index (Pompe-MDPI) was developed and validated as a sensitive biomarker of disease progression. Longitudinal analysis demonstrated that Pompe-MDPI captured gait deterioration over one year, outperforming traditional measures like the six-minute walk test in sensitivity. These findings support the use of wearable gait analysis as a clinically meaningful, scalable tool for monitoring motor function in LOPD, with implications for both patient care and therapeutic trials.