Purpose Balance disorders increase fall risk in older adults, yet global gait parameters may lack sensitivity to detect subtle impairments. This study evaluates whether detailed spatiotemporal and plantar-pressure measures more effectively identify balance-related gait alterations. Methods Fifty-two older adults (55–82 years) were assigned to a balance disorders group (n = 26) or a control group (n = 26) based on fall history and Dizziness Handicap Inventory scores. Barefoot walking trials were recorded using a pedobarographic platform. Scalar spatiotemporal gait parameters and pressure-related signals were normalized to body height and body mass, respectively. Dynamic foot balance parameters were analyzed as time-normalized waveforms over the stance phase. Between-group and within-group comparisons were performed using conventional statistical methods and Statistical Parametric Mapping (SPM). Results No statistically significant between-group differences were found in global spatiotemporal gait parameters or in group-averaged mean and variability waveforms of dynamic foot balance measures. However, within-group SPM analyses presented stance-phase intervals with significant left–right asymmetries in selected dynamic foot balance parameters, occurring more frequently in participants with balance disorders. These asymmetries were parameter- as well as phase-specific and showed substantial inter-individual variability. Conclusions Older adults with balance disorders may preserve overall gait characteristics during steady-state walking while exhibiting subtle, localized asymmetries in dynamic foot balance. Waveform-based plantar pressure analysis may capture balance-related gait alterations not detected by global gait measures, although the clinical relevance of these findings requires further investigation.
Purpose The aim of this study was to assess the effect of completing a water rescue course on lower limb power parameters assessed using the countermovement jump test (CMJ) and on performance on task-based tests in the water. Methods Nineteen female lifeguard candidates (aged 18.68 ± 5.89 years) participated in the study. A repeated measures design was used, with assessments conducted before (PRE) and after course completion (POST). The following CMJ variables were analyzed: jump height, push-off force, landing force, maximum concentric power, maximum speed, and take-off velocity. Functional performance in the water was assessed using tests in the 50m, freestyle, front crawl, breaststroke, and backstroke. Results After completing the course, significant improvements were observed in selected CMJ parameters, including jump height (+8.47%; p = 0.008), rebound force (+23.42%; p = 0.006), maximum concentric power (+8.72%; p = 0.047), maximum velocity (+2.81%; p = 0.017), and rebound speed (+3.77%; p = 0.022). In the water rescue trials, significant reductions in completion times were observed for the front crawl (-4.23%; p = 0.013), 50m kick (-5.48%; p < 0.001), two-hand pull (-7.30%; p = 0.004), one-handed chin pull (-5.40%; p < 0.001), and chest pull (-3.89%; p = 0.006). Conclusions Participation in a water rescue course significantly improves both lower limb strength and functional performance in the water. However, transfer of lower limb strength from land-based tasks to rescue tasks is moderate, and performance in the water depends more on specific functional adaptations and task-specific skills than on limb strength.
Purpose The sit to walk (STW) transition serves not merely as a highly prevalent functional movement in activities of daily living, but also as a pivotal task paradigm for elucidating overall motor control strategies. The objective of this study was to comparatively evaluate the stability control characteristics between two groups during STW transitions, and to explore the underlying mechanisms associated with potential fall risks. Methods Lower-limb differences between the two groups were assessed using independent samples t-tests, Statistical Parametric Mapping (SPM), and phase-average comparisons, supplemented by correlation analyses. Results Peak horizontal momentum (HM) showed significant differences between the two groups across the entire phase (p < 0.05). The anteroposterior margin of stability (AP MOS) differed significantly during the extension and swing phases (p < 0.03), whereas knee range of motion (ROM) and center of pressure (COP) excursion velocity exhibited significant differences only in the unloading phase (p < 0.05). In the young group, the margin of stability, HM, and joint kinematics demonstrated significant strong negative correlations (r<-0.5, p<0.05), whereas these correlations were generally attenuated in the older adult group. Conclusions Older adults tend to adopt a conservative postural control strategy during STW transitions, prioritizing enhanced stability by restricting forward HM and prolonging movement duration. These findings elucidate the characteristics underlying fall risks during STW in older adults, providing a scientific rationale for the design of targeted biomechanical interventions.
Purpose This study presents a comparative technical evaluation of two implemented electrocardiogram (ECG) acquisition paths based on AD8232 and MAX30003 under controlled laboratory conditions. The analog front-end paths were designed in accordance with the reference circuits and design recommendations. Methods Synthetic ECG signals were generated using a patient simulator for heart rates of 30, 60, 80, and 100 BPM and were recorded simultaneously from both acquisition paths at a sampling frequency of 128 Hz. The processing pipeline comprised time-axis standardization, baseline correction, local cross-correlation-based synchronization, linear amplitude alignment, QRS detection using a simplified Pan-Tompkins-inspired procedure, and median-centered amplitude standardization. Signal similarity after preprocessing was assessed using time-domain metrics, including Pearson correlation, MAE, RMSE, and NRMSE. In addition, spectral analysis based on Welch’s method was performed using absolute bandpower values and relative bandpower indices. Results After synchronization and normalization, both acquisition paths showed very high morphological agreement. Error metrics remained low and comparable between recordings. Spectral analysis revealed marked differences in absolute power, reflecting properties of the complete acquisition chains rather than signal quality alone. Relative spectral indices indicated lower 50 Hz interference and lower high-frequency content in the MAX30003-based path. Conclusions Under the tested synthetic and controlled conditions, both implemented ECG acquisition paths provided comparable post-processed morphological representations of the ECG signal, while differing in spectral characteristics. The MAX30003-based path showed lower relative out-of-band spectral content. However, these findings should be interpreted as implementation-specific technical observations rather than evidence of general device superiority. Further validation in human recordings and real-world conditions is required.
Purpose Growth-guidance spinal systems are used in children with progressive spinal deformity to reduce repeated surgical lengthening while allowing continued spinal growth. Their long-term biomechanical behaviour under cyclic loading remains insufficiently characterised. This study assessed whether diamond-like carbon coating of rods and sliding screws affects axial stiffness and local kinematics in a growth-guidance spinal system. Methods Twelve porcine Th11–L7 spine specimens were assigned to three groups: non-instrumented control, non-coated titanium alloy instrumentation, and instrumentation with diamond-like carbon-coated components. Specimens underwent 100,000 cycles of axial compression. Axial stiffness was evaluated after 100 and 100,000 cycles. Vertebral range of motion from Th12 to L6 was measured using rigid body markers, and local rod–screw interface displacements were assessed using three-dimensional digital image correlation. Results Stiffness increased similarly in all groups up to approximately 80,000 cycles. After 100,000 cycles, construct stiffness was approximately 6% lower in diamond-like carbon coated instrumentation than in non-coated titanium alloy instrumentation. In the coated construct, segmental mobility at Th12–L1 and L5–L6 remained closer to the non-instrumented condition, whereas non-coated titanium alloy instrumentation showed reduced motion at these levels. Digital image correlation demonstrated time-dependent differences in local rod–screw interface kinematics between coated and non-coated constructs. Conclusions Diamond-like carbon coating did not increase overall construct stiffness during cyclic axial compression, but it modified local mechanical behaviour and was associated with preservation of segmental mobility at selected levels. Further studies should determine whether these effects are maintained under multi-axis loading and in long-term tribological conditions.
Purpose Existing brain injury criteria mainly focus on assessing injuries in a single collision and lack of assessment of injuries in real-world multiple collisions. Methods Based on the head kinematics signals of rats in repeated collisions and pedestrian accident data, the frequency domain characteristics of linear and rotational motion signals were analyzed by wavelet packet transform. Finally, the ROC curve was used to evaluate the effectiveness of injury index prediction. Results The results of animal experiments show that as the impact intensity increases, the energy in the frequency domain during the rotational motion increases accordingly. In linear motion, there is a threshold of impact strength. The prediction results show that the prediction effect of metrics based on frequency domain features is better than that of traditional brain injury criteria. Conclusions The new metrics based on wavelet packet energy accumulation can better predict the injury in multiple head impact scenarios in pedestrian collisions. When combined with the acceleration response signal's amplitude-frequency accumulation characteristics, they can be used as a potential method to assess repetitive head injury.
Purpose This study aimed to determine the acute impact of employing expert modeling on tennis serve performance in 7–9-year old boys. Methods Seventeen players undertook three conditions sequentially: control condition (traditional training), video modeling condition 1, and condition 2. Each participant performed 16 serves per condition (8 per diagonal), with one week between each condition. Serve performance outcomes were evaluated under each condition. Results Repeated measures ANOVAs revealed significant main effects for conditions (F(2, 32) = 85.61, p < 0.001, η²p = 0.843) and point areas (F(2, 32) = 274.76, p < 0.001, η²p = 0.945), with a significant interaction (F(4, 64) = 7.08, p < 0.001, η²p = 0.307). Both video modeling conditions outperformed the control condition, with VMC2 showing the greatest benefits for successful serve (F(2, 32) = 85.61, p < 0.001, η²p = 0.843) and ball impact (F(2, 32) = 70.29, p < 0.001, η²p = 0.815), while no effect was observed for ball speed (p = 0.116–0.458). Significant condition × test interactions were observed for point areas 1, 3, and 5 (p < 0.001, η²p = 0.614–0.720). Pairwise comparisons with Bonferroni correction revealed significant differences between all conditions (p < 0.001), with VMC2 producing the largest gains across all point areas. Conclusions Our findings support that the adoption of expert video modeling for coaching/learning tennis serves to expedite the process of technique refinement and facilitate performance enhancement.
Purpose Purpose: Center Of Pressure (COP) is key to understand dynamic stability of Gait Initiation (GI) and Gait Termination (GT). This study investigated the COP between GI and GT from discrete parameters and time-varying coordinates. Methods Methods: Forty healthy male adults participated in this study while walking on a pathway facilitated a Novel EMED pressure plate to record the COP trajectory during the initiation and termination. Discrete parameters of COP duration, velocity, and coordinate deviation were compared, and continuous mediolateral and anteroposterior trajectories were analyzed using SPM1d (one-dimensional Statistical Parametric Mapping) and fPCA (functional Principal Component Analysis). Results Results: GT with left foot had shorter Forefoot Contact Phase (FFCP) and longer Flatfoot Phase (FFP) than GI, suggesting a quick transition to full-foot support to maintain dynamic stability. COP velocity during Initial Contact Phase (ICP) increased significantly during GT, which may be a strategy to counteract instability. GT also showed larger coordinate deviation during ICP, especially in the mediolateral coordinate, from the SPM analysis. The fPCA further reported the variations of anteroposterior and mediolateral coordinates, serving as the reference of COP profiles for GI and GT analysis in the dominant (right) and non-dominant (left) foot. Conclusions Conclusions: Knowledge of discrete parameters and continuous coordinates in the COP trajectory from this study may provide implications to reduce inter-limb asymmetry during gait, and may serve as reference to assess gait disorders and rehabilitation.
Purpose The study aimed to investigate whether saxophonists experience muscle tone variations in the masseter muscles and the descending parts of the trapezius dorsi muscle in resting conditions and during activities that load the stomatognathic system. Methods This study included 30 men aged 20-30 years, students of saxophone classes at the Jazz and Popular Music Department of randomly selected Music Academies in Poland (average saxophone playing experience: 12,57±2,06 years). Muscle tension of selected masticatory and back muscles was evaluated both at rest and during activities taxing the stomatognathic system using a Noraxon electromyograph. Friedman's repeated measures analysis of variance for ranks (ANOVA) with Dunn's post-hoc test was used to analyze the results. Results Statistically significant differences were found in the maximal contraction amplitude values of the masseter muscle obtained in subsequent test sequences, both on the right (p<0.001) and left (p<0.001) sides, as well as in the descending portion of the trapezius muscle on the right (p<0.001) and left (p<0.001) sides. Statistically significant differences were noted in the maximal contraction amplitude values of the analysed muscles between the individual test phases. Conclusions In saxophonists, variation occurs in the tension of the masseter muscles and the descending parts of the trapezius muscle during rest and activity, which burdens the stomatognathic system The tension in these muscles tends to increase during free opening and closing of the mouth and when clenching the teeth on the mouthpiece. It reaches its highest values while playing the instrument, and after ceasing activity, it decreases during the rest phase.
Purpose This study aims to explore the application value of multi-parameter surface electromyography (sEMG) combined with Motor Status Scale (MSS) score in the evaluation of post-stroke upper limb spasticity. Methods 42 patients with post-stroke upper limb spasticity were enrolled in a self-controlled study design. All patients received routine neurological treatment and rehabilitation training. The iEMG, root mean square (RMS) value, median frequency (MF), and mean power frequency (MPF) were analyzed, and the co-contraction ratio (CR) was calculated. Results After treatment, the iEMG, RMS, MF, and MPF values of the forearm flexor and extensor muscle groups significantly increased in the 42 patients (P<0.05). At rest, the RMS and iEMG values of the affected forearm flexor and extensor muscles were significantly higher than those of the unaffected side and the healthy control group (P<0.05). During active movement, the RMS and iEMG values of the affected muscles were significantly lower than those of the unaffected side and the healthy control group (P<0.05). The total MSS score was significantly increased compared with that before treatment (P<0.001). sEMG parameters were negatively correlated with balance function parameters (P<0.05). The modified Ashworth Scale assessment showed that the total effective rate was 95.24%. Receiver operating characteristic curve analysis revealed that the AUC values of iEMG, RMS, MF, and MPF for evaluating the efficacy were 0.748, 0.731, 0.737, and 0.826, respectively (P<0.05). Conclusions Multi-parameter sEMG analysis combined with MSS score can objectively and quantitatively evaluate the functional status of post-stroke upper limb spasticity.
Purpose This study aimed to investigate the acute effects of loaded whole-body vibration (WBV) training on both the spatial (muscle weightings) and temporal (activation patterns) characteristics of lower-limb muscle synergies during running. Methods Thirty-nine healthy male recreational runners completed running biomechanics assessments before and after a single loaded whole-body vibration (WBV) training session. The intervention comprised dynamic squats on a vibration platform (20 Hz, 6 mm) while wearing a weighted vest. Surface EMG signals from lower-limb muscles were recorded and decomposed into muscle synergies using non-negative matrix factorization. Pre- and post-intervention comparisons were conducted using paired-sample t-tests or Wilcoxon signed-rank tests. Results Muscle synergy analysis consistently identified five synergies before and after the intervention. Compared with pre-intervention, muscle weights increased for medial gastrocnemius in SYN1 (p = 0.036), peroneus longus in SYN3 (p = 0.045), rectus femoris (p = 0.007), and vastus lateralis (p = 0.049) in SYN5, while those for the vastus medialis in SYN2 (p = 0.013) and vastus lateralis in SYN4 (p = 0.031) decreased. After the intervention, temporal parameters demonstrated a longer activation duration in SYN1 (p = 0.028), as well as earlier onset (p = 0.047), earlier offset (p = 0.033), and earlier peak activation (p = 0.041) in SYN3. Conclusions A single-loaded WBV session acutely reorganized muscle synergy, indicating rapid neural adaptations that may enhance propulsion efficiency and landing stability. Overall, these enhancements support the use of single-loaded WBV session as an effective warm-up strategy for running.
Purpose Gait after a stroke is characterised by a number of abnormalities, such as reduced walking speed, impaired stability, as well as asymmetry. This study presents a novel assistive device designed to aid gait training in stroke survivors by providing integrated visual, auditory, and tactile stimuli informing about lower limb weight-bearing and step length symmetry. Methods Nine patients with hemiparesis at an early stage post-stroke were included in the study. A markerless 3D system was applied to assess gait. Participants walked at self-selected pace and were wearing their own orthotic devices. Gait speed, step length, gait phase duration, and cadence were measured. A symmetry index was calculated for step length and lower limb weight-bearing. The programme consisted of 10 gait training sessions with the assistive device, each session taking 30 minutes. Results Improved gait parameters were observed when acoustic and visual prompting was applied alone or in combination with tactile prompting. At the end of the programme, decrease in step length asymmetry was observed and there was no statistically significant difference in the mean step length between the paretic and non-paretic lower limb (p=0,002). An improvement was shown in the distribution of lower limb weight-bearing in a standing position, and the mean value of limb weigh-bearing symmetry index was significantly lower compared to the baseline (p=0,0001). Conclusions Integrated tactile, auditory, and visual prompting may beneficially affect post-stroke gait rehabilitation, by inducing improvements in gait pattern, symmetry, body balance, and walking speed.
Purpose To investigate the effects of an 8-week core-strength training program on smash accuracy in female badminton athletes. Methods Forty female badminton players were randomly assigned to an experimental group (EG) or a control group (CG). Both groups continued regular badminton training, while the EG additionally completed an 8-week structured core-strength program. Smash accuracy was assessed under three conditions—stationary, jumping, and moving smashes—at baseline, mid-intervention (posttest 1), and post-intervention (posttest 2). Data were analyzed using generalized estimating equations (GEE) to examine group effects, time effects, and group × time interactions, with age, training experience, height, and body mass included as covariates. Results GEE analysis revealed significant group, time, and group × time interaction effects for in-situ, jumping, and moving smash accuracy (p < 0.05). Post hoc analyses showed that the EG demonstrated significant improvements in smash accuracy from baseline to posttest 2 across all three smash conditions, whereas improvements in the CG were limited or non-significant. Effect size analysis further indicated that the magnitude of improvement in smash accuracy was substantially greater in the EG than in the CG. None of the covariates showed a significant influence on smash accuracy outcomes. Conclusions An 8-week core-strength training program significantly enhances in-situ, jumping, and moving smash accuracy in female badminton athletes. These findings highlight the importance of core stability and strength in optimizing technical performance and suggest that systematic core training should be incorporated into routine badminton training programs to improve smash accuracy and overall competitive performance.
Purpose The main objective of study was to analyse gait kinematics in terms of symmetry among groups of children with hemiplegic cerebral palsy (CP). Methods The study was conducted using the BTS SMART three-dimensional (3D) gait analysis optoelectronic system. Passive reflective markers were placed on subjects' bodies in accordance with the Davis protocol. The study involved 43 children, aged seven years. Group 1 (n=18) comprised patients not using any orthopaedic equipment. Group 2 (n=14) consisted of children using a unilateral ankle-foot orthosis on the affected side. Group 3 (n=11) included children using bilateral ankle-foot orthoses. Results Improvement in gait symmetry was observed in orthosed children, and the majority of changes were significant (p<0.05). Thus, it appears that this type of orthopaedic equipment affects not only the ankle joint, but also higher levels of the biokinematic chain. Conclusions In accordance with the results, it was shown that in patients using bilateral ankle-foot orthoses, a more positive effect of the described phenomena could be expected.
Purpose To conduct a comparative analysis of the trajectory and joint torque differences between exoskeleton-type and end-effector-type upper limb rehabilitation training mechanisms during rehabilitation movements, and to identify the more suitable mechanism type for passive upper limb rehabilitation training. Methods Based on the Brunnstrom rehabilitation approach, upper limb rehabilitation movements for the passive training phase were selected. Muscle force simulations were conducted using OpenSim software to identify movements with high muscular engagement. The structures of the two mechanisms were designed through degree-of-freedom analysis, and motion trajectory simulations were completed in ADAMS. In MATLAB, discrete Fréchet distance is used to evaluate trajectory similarity. Calculate the error of comparing the trajectories of two mechanisms with the trajectory of the human body model at the same time using R2. In ADAMS, compare the rotational torque changes of the shoulder, elbow, and wrist joints between two mechanisms. Results The movement trajectory of the exoskeleton mechanism closely matched the natural motion of the human upper limb, error ≤ 2 mm, trajectory R2 ≥ 0.9975. Its joint torque distribution was balanced, with a maximum torque ≤ 40 Nm. The trajectory error of the end-effector mechanism ≤5 mm, trajectory R2 ≥ 0.9910 and torques were concentrated at the wrist and elbow joints, some values ≥ 40 Nm. Conclusions The exoskeleton-type upper limb rehabilitation training mechanism demonstrates superior performance in terms of movement trajectory adaptability and joint protection compared to the end-effector-type mechanism, making it more suitable for passive movement assistance and joint range of motion recovery training.
Purpose Birmingham Hip Resurfacing requires precise determination of the femoral head-neck axis to minimize risks of neck fracture, cortical notching, and component malposition. Existing methods rely on intraoperative estimation, generic guides, or semi-manual planning, all susceptible to anatomical variability. This study presents an automated parametric algorithm for determining the femoral neck axis from CT-derived models to support patient-specific surgical planning. Methods The method utilizes segmented CT data converted to STEP format. Following artifact removal, the algorithm performs iterative cross-sectional analysis along the neck axis, identifying characteristic anatomical points: the neck isthmus, head centroid, and head apex. The axis is approximated from these points and refined through rotational optimization in 5 degree and 0.5 degree increments. The algorithm was evaluated on four femoral models in five spatial orientations each (20 test cases). Results The algorithm achieved a mean validation score of 4.1/5.0 (82%). Three models with high-quality meshes scored 4-5, while one model with extensive artifacts scored 1-2, confirming sensitivity to input data quality. Repeatability was plus/minus 0.5 degrees in angular orientation and plus/minus 0.2 mm in isthmus position. Error propagation analysis indicated that CT-induced uncertainty (approximately 1.6 degrees upper bound) constitutes the limiting factor, not algorithmic precision. Physical prototyping confirmed viability for patient-specific instrument design. Conclusions The method provides reproducible, automated femoral neck axis determination with plus/minus 0.5 degrees repeatability -- approximately three times smaller than CT-induced uncertainty and an order of magnitude below the 5-7 degree errors of conventional jigs. Future work should validate the method against clinical outcomes in larger cohorts.
Purpose The aim of this study was to determine the required concentration of magnetic nanoparticles to achieve the therapeutic temperature in prostate cancer treatment during magnetic hyperthermia, taking into account the influence of the geometry of the heated region. In this context four types of magnetic nanoparticles were considered in the calculations: maghemite, magnetite, cobalt ferrite and barium ferrite. Methods An air coil generating an alternating magnetic field was designed to obtain a uniform magnetic field distribution in the prostate region. The magnetic field distribution was used to calculate the volumetric power density generated by the magnetic nanoparticles. Thermal simulations were conducted for a naturalistic prostate tumour and three geometric primitives (cylinder, sphere and cube) placed in the same anatomical location for comparison purposes. To evaluate the concentration of various types of magnetic nanoparticles, the Linear Response Theory was used. Results The concentrations of magnetic nanoparticles for the tumour, cylinder, sphere and cube required to achieve the therapeutic temperature were determined. The results revealed noticeable differences between the tumour model and simplified geometries in terms of both required nanoparticle concentration and heating efficiency. Conclusions The obtained results indicate that the concentration of magnetic nanoparticles required to achieve therapeutic temperatures strongly depends on the geometry of the heated region, type of magnetic nanoparticles and their concentration. Furthermore, the use of simplified geometrical models may lead to significant inaccuracies in predicting heat generation, suggesting that realistic tumour geometry should be considered in magnetic hyperthermia studies. These findings highlight the necessity of individualized treatment planning.
Purpose This study investigated the effects of lower visual field (LVF) obstruction on adaptive gait during obstacle crossing in healthy young adults using a novel "box-carrying obstruction method. Methods Fifteen male participants (age, 20.9 ± 1.3 years; height, 178.6 ± 5.0 cm; weight, 71.7 ± 7.5 kg; leg length, 91.7±5.1 cm) walked with or without a box (30 × 30× 25 cm or 30 × 30× 35 cm) while crossing obstacles at heights of 10%, 20%, and 30% of leg length in randomized order. Measured gait parameters included step length, step width, step velocity, penultimate foot placement (PFP), toe distance (TD), heel distance (HD), and lead/trail limb vertical toe clearance (LTC/TTC). Statistical analyses included repeated-measures ANOVA and Pearson correlations. Results No significant interactions between LVF obstruction and obstacle height were observed (all P > 0.05). Obstacle height significantly reduced step velocity (P = 0.002), while LVF obstruction independently affected TD (P < 0.001), HD (P < 0.001), LTC (P < 0.001), and TTC (P < 0.001). Negative correlations were found between TD and HD (r = -0.46; P < 0.001), and positive correlations between LTC and TTC (r = 0.6; P < 0.001). Conclusions LVF obstruction had a broader impact on spatial gait adjustments than obstacle height, with step velocity serving as the primary adaptive strategy for higher obstacles. The strong coupling between lead and trail limb clearances suggests interlimb coordination, highlighting the role of shared sensorimotor information in obstacle negotiation.
Purpose The Up and Go test is a widely used tool among physical therapists for assessing functional mobility. Advances in inertial measurement unit sensors (IMU) enable the analysis not only of total walk duration, but also of the duration of individual walk phases, as well as the determination of spatiotemporal and kinematic parameters for each phase. Consequently, the IMU-supported version of the test up and go, referred to as the instrumented up and go, enables a substantially more comprehensive analysis of locomotor patterns. The aim of this study was to evaluate the extent to which spatiotemporal and kinematic parameters obtained during the instrumented timed up nad go test correspond to those observed during walking in older adults. Methods The study included 30 adults aged over 60 years, who performed the instrumented up and go test as well as a 10-m walk test at a self-selected walking speed. The study involved the analysis of selected spatiotemporal parameters, including cadence, velocity, stride length, and the percentage contribution of gait phases, as well as analysis of pelvic tilt and lower-limb kinematics, including ranges of motion. Results The results indicate that while the instrumented up and go test does not directly reflect natural walk patterns, selected parameters may still provide complementary information relevant to functional mobility assessment in older adults. Conclusions The instrumented up and go test should not be considered a substitute for standard walk analysis.
Purpose Reliable decoding of lower-limb motor intent from surface electromyography signals remains challenging due to electrode placement variability, muscle fatigue, inter-subject differences, and differences between able-bodied and amputee populations. This study evaluates the robustness of commonly used machine-learning approaches for lower-limb electromyography classification under these sources of variability. Methods A unified benchmarking analysis was conducted across four lower-limb electromyography datasets: UCI Lower Limb Electromyography, Vastus Medialis Electromyography, K2MUSE, and the Above-Knee Sit-to-Stand amputee dataset. Three classification models—support vector machine, random forest, and multilayer perceptron—were evaluated using standardized preprocessing and feature extraction pipelines. Cross-subject robustness was assessed using leave-one-subject-out validation on the K2MUSE dataset. A preliminary transfer-learning experiment was also performed by fine-tuning a multilayer perceptron pretrained on able-bodied data using the amputee sit-to-stand dataset. Results The random forest model achieved the highest performance in datasets with lower variability, with macro F1 scores of 0.9033 ± 0.0048 on the UCI dataset and 0.6000 ± 0.0087 on the Vastus dataset. Under cross-subject non-ideal acquisition conditions, performance decreased for all models. The multilayer perceptron showed competitive robustness in the K2MUSE Ideal-versus-Fatigue LOSO evaluation (0.476 ± 0.058 macro-F1), while random forest showed the most compact subject-wise distribution (median 0.476 [IQR 0.049]). In the transfer-learning experiment, fine-tuning improved performance on the Above-Knee STS dataset relative to training from scratch (macro-F1 0.5540 ± 0.0141 vs. 0.5203 ± 0.0371). Conclusions These results highlight the different strengths of classical and neural machine-learning models under varying electromyography variability conditions and provide a foundation for more robust intention-decoding systems for adaptive lower-limb prosthetic control.
