Background: Three-dimensional gait analysis, supported by advanced sensor systems, is a crucial component in the rehabilitation assessment of post-stroke hemiplegic patients. However, the sensor data generated from such analyses are often complex and challenging to interpret in clinical practice, requiring significant time and complicated procedures. The Gait Deviation Index (GDI) serves as a simplified metric for quantifying the severity of pathological gait. Although isokinetic dynamometry, utilizing sophisticated sensors, is widely employed in muscle function assessment and rehabilitation, its application in gait analysis remains underexplored. Objective: This study aims to investigate the use of sensor-acquired isokinetic muscle strength data, combined with machine learning techniques, to predict the GDI in hemiplegic patients. This study utilizes data captured from sensors embedded in the Biodex dynamometry system and the Vicon 3D motion capture system, highlighting the integration of sensor technology in clinical gait analysis. Methods: This study was a cross-sectional, observational study that included a cohort of 150 post-stroke hemiplegic patients. The sensor data included measurements such as peak torque, peak torque/body weight, maximum work of repeated actions, coefficient of variation, average power, total work, acceleration time, deceleration time, range of motion, and average peak torque for both flexor and extensor muscles on the affected side at three angular velocities (60°/s, 90°/s, and 120°/s) using the Biodex System 4 Pro. The GDI was calculated using data from a Vicon 3D motion capture system. This study employed four machine learning models—Lasso Regression, Random Forest (RF), Support Vector regression (SVR), and BP Neural Network—to model and validate the sensor data. Model performance was evaluated using mean squared error (MSE), the coefficient of determination (R²), and mean absolute error (MAE). SHapley Additive exPlanations (SHAP) analysis was used to enhance model interpretability. Results: The RF model outperformed others in predicting GDI, with an MSE of 16.18, an R² of 0.89, and an MAE of 2.99. In contrast, the Lasso Regression model yielded an MSE of 22.29, an R² of 0.85, and an MAE of 3.71. The SVR model had an MSE of 31.58, an R² of 0.82, and an MAE of 7.68, while the BP Neural Network model exhibited the poorest performance with an MSE of 50.38, an R² of 0.79, and an MAE of 9.59. SHAP analysis identified the maximum work of repeated actions of the extensor muscles at 60°/s and 120°/s as the most critical sensor-derived features for predicting GDI, underscoring the importance of muscle strength metrics at varying speeds in rehabilitation assessments. Conclusions: This study highlights the potential of integrating advanced sensor technology with machine learning techniques in the analysis of complex clinical data. The developed GDI prediction model, based on sensor-acquired isokinetic dynamometry data, offers a novel, streamlined, and effective tool for assessing rehabilitation progress in post-stroke hemiplegic patients, with promising implications for broader clinical application.

Objective: The classic excitation/inhibition dichotomy may be insufficient to describe rTMS mechanisms in the aging brain. This study investigated immediate whole-brain resting-state functional connectivity effects of 10 Hz (high-frequency) and 1 Hz (low-frequency) rTMS over the left dorsolateral prefrontal cortex (DLPFC) in healthy older adults. Methods: Thirty healthy older adults (aged 60–75 years) participated in a randomized, single-blind, crossover study, and underwent 20-min 10 Hz and 1 Hz rTMS in separate visits. A 106-channel fNIRS system was used to record resting-state activity before and immediately after each intervention. Functional connectivity was analyzed at the channel, region-of-interest (ROI) and network summary levels, including graph-theoretic metrics and distance-stratified connectivity summaries. Results: At the network summary level, 10 Hz stimulation was associated with relatively more positive changes in global topology and spatially distributed connectivity summaries, whereas 1 Hz stimulation showed the opposite overall trend. In the graph-theoretic analyses, stimulation frequency × time interaction effects were observed for global efficiency, local efficiency, clustering coefficient, and mean node strength. At the edge level, only a small number of effects survived FDR correction, and the broader connection-wise patterns were therefore interpreted as exploratory. Uncorrected analyses suggested widespread enhancement after 10 Hz stimulation and widespread reduction after 1 Hz stimulation, together with localized paradoxical effects, including selective decreases after 10 Hz and selective increases after 1 Hz (e.g., bilateral primary motor cortex connectivity). Conclusions: These findings suggest that 10 Hz and 1 Hz rTMS over the left DLPFC are associated with different patterns of immediate whole-brain network reconfiguration in healthy older adults. The presence of localized paradoxical effects further suggests that rTMS responses in the aging brain may involve more complex forms of reorganization than a simple excitatory/inhibitory dichotomy would predict. Significance: The present study provides preliminary support for a network-level perspective on neuromodulation in older adults and highlights the value of whole-brain fNIRS for characterizing distributed responses to rTMS. Larger, sham-controlled, behavior-linked, and longitudinal studies are needed to determine the robustness and functional significance of these effects.

Background: Three-dimensional gait analysis plays a crucial role in the rehabilitation assessment of post-stroke hemiplegic patients. However, the data generated from such analysis is often complex and difficult to interpret in clinical practice, requiring extensive time and complicated procedures. The Gait Deviation Index (GDI) serves as a simplified metric for quantifying the severity of pathological gait. Although isokinetic dynamometry is widely employed in muscle function assessment and rehabilitation, its application in gait analysis remains underexplored. Objective: This study aims to investigate the use of isokinetic muscle strength data obtained from the Biodex system, combined with machine learning techniques, to predict the GDI in hemiplegic patients. Methods: A cohort of 150 post-stroke hemiplegic patients was evaluated. The peak torque, peak torque/body weight, maximum work of repeated actions, coefficient of variation, average power, total work, acceleration time, deceleration time, range of motion, and average peak torque for both flexor and extensor muscles on the affected side were measured at three angular velocities (60°/s, 90°/s, and 120°/s) using the Biodex System 4 Pro. GDI was calculated using data from a Vicon three-dimensional motion capture system. The study employed four machine learning models—Lasso regression, random forest (RF), support vector regression (SVR), and BP neural networks—to model and validate the data. Model performance was evaluated using mean squared error (MSE), the coefficient of determination (R²), and mean absolute error (MAE). SHapley Additive exPlanations (SHAP) analysis was used to enhance model interpretability. Results: The RF model outperformed others in predicting GDI, with an MSE of 16.18, an R² of 0.89, and an MAE of 2.99. In contrast, the Lasso regression model yielded an MSE of 22.29, an R² of 0.85, and an MAE of 3.71. The SVR model had an MSE of 31.58, an R² of 0.82, and an MAE of 7.68, while the BP neural network exhibited the poorest performance with an MSE of 50.38, an R² of 0.79, and an MAE of 9.59. SHAP analysis identified the maximum work of repeated actions of the extensor muscles at 60°/s and 120°/s as the most critical features for predicting GDI, underscoring the importance of muscle strength metrics at varying speeds in rehabilitation assessments. Conclusion: This study highlights the potential of machine learning techniques in analyzing complex clinical data. The developed GDI prediction model, based on isokinetic dynamometry data, offers a novel, streamlined, and effective tool for assessing rehabilitation progress in post-stroke hemiplegic patients, with promising implications for broader clinical application.

Background: To investigate whether the common clinical practice of setting transcranial magnetic stimulation (TMS) parameters by scaling from the motor threshold (MT) measured at the primary motor cortex (M1) can be expected to yield comparable stimulation at complex brain regions such as the dorsolateral prefrontal cortex (DLPFC). Methods: Personalized head models were constructed from T1- and T2-weighted magnetic resonance imaging (MRI) scans from 20 healthy elderly adults from a public Chinese dataset. SimNIBS 4.1 was used for tissue segmentation, mesh generation, and TMS electric field (E-field) simulations. A MagVenture MC-B70 figure-of-eight coil was used to simulate stimulation over the M1 and DLPFC (C3/F3, 10–10 electroencephalography (EEG) system). First, both targets were simulated at a fixed intensity of 60% of maximum stimulator output (MSO) to isolate anatomical contributions. Second, an MT-equivalent reference output framework was implemented by calibrating subject-specific output to an M1 reference E-field level of 60 mV/mm, using two calibration tracks (Mean-based and Max-based), and the DLPFC was evaluated at 80–120% of the subject-specific reference output. Results: Under fixed-output stimulation, the mean and robust maximum E-field magnitudes on the gray matter surface and within gray matter volume were significantly lower in the DLPFC region of interest (ROI) than in the M1 ROI (PFDR < 0.05). The normal (surface-perpendicular) component did not differ significantly between targets (PFDR > 0.05), whereas the tangential (surface-parallel) component was significantly weaker at the DLPFC (PFDR < 0.05). Under MT-equivalent scaling, increasing intensity reduced the overall magnitude gap, with approximate parity around 110% of the reference output, but a component trade-off persisted: scaling that brought tangential fields closer to the M1 reference tended to increase normal fields beyond the M1 reference. Conclusion: Fixed-output simulations show target-wise differences in both E-field magnitude and component balance between the M1 and DLPFC. An M1-referenced, MT-equivalent scaling framework can reduce magnitude differences but does not necessarily align component balance at the DLPFC. These findings support target-specific, E-field–informed parameter selection and should be tested in prospective studies linking modeled E-fields to experimental readouts and clinical outcomes.