Development of a convolutional neural network (CNN) based assessment exercise recommendation system for individuals with chronic stroke: a feasibility study.

BACKGROUND: The use of artificial intelligence (AI) is revolutionizing nearly every aspect of healthcare, but the application of AI in rehabilitation is lagging behind. Clinically, gait parameters and patterns are used to evaluate stroke-specific impairment. We hypothesized that gait kinematics of individuals with stroke provide rich information for the deep-learning to predict the clinical decisions made by physiotherapist. OBJECTIVE: To investigate whether the results of clinical assessments and exercise recommendations by physiotherapists can be accurately predicted using a deep-learning algorithm with gait kinematics data. METHOD: In this cross-sectional study, 40 individuals with stroke were assessed by a physiotherapist using the lower-extremity subscale of the Fugl-Meyer Assessment (FMA-LE) and Berg Balance Scale (BBS). The physiotherapist also decided whether or not the single-leg-stance was an appropriate balance training for each participant. The participants were classified as having good mobility and a low fall risk based on the cutoff scores of the two clinical scales. A convolutional neural network (CNN) was trained using gait kinematics to predict the assessment results and exercise recommendations. RESULTS: The trained model accurately predicted the results of the clinical assessments and decisions with an average prediction accuracy of 0.84 for the FMA-LE, 0.66 for the BBS, and 0.78 for the recommendation of the single-leg-stance exercise. CONCLUSIONS: This CNN deep-learning model provided time-effective and accurate prediction of clinical assessment results and exercise recommendations. This study provides preliminary evidence to support the use of biomechanical data and AI to assist treatment planning and shorten the decision-making process in rehabilitation.

Incidence, Mechanisms, and Characteristics of Injuries in Pole Dancers: A Prospective Cohort Study.

OBJECTIVE: Pole dancing is a challenging physical activity. Prospective injury studies in pole dancing are lacking. The aim of this study was to describe the incidence, mechanisms, and characteristics of injuries in pole dancers. METHODS: A total of 66 pole dancers from 41 studios across Australia were prospectively followed over 12 months. An intake questionnaire was administered including items on pole dancers' demographics and training characteristics. Exposure was assessed using a daily online training diary. Self-reported injury data were collected via an incident report form and subsequently coded using the Orchard Sports Injury Classification System. Injuries occurring during pole-specific and pole-related activities were included in the analyses. RESULTS: The sample included 63 females and 3 males, mean age 32.3 ± 8.9 years and mean pole training experience 3.5 ± 2.8 years. 25 of 66 participants completed the full study. The 1-year incidence of all new injuries was 8.95/1,000 exposure hours (95% CI 6.94 - 10.96), 7.65/1,000 hrs (95% CI 5.79 - 9.51) for pole-specific injuries and 1.29/1,000 hrs (95% CI 0.53 - 2.06) for pole-related injuries. A total of 103 injuries occurred, 62.1% of which were sudden onset and 37.9% gradual onset. Mechanism of onset included 54.4% acute and 45.6% repetitive in nature. Shoulder (20.4%) and thigh (11.7%, majority ham¬string) were the most reported anatomic injury sites. Non-contact mechanisms accounted for the majority of injuries (57.3%). The most reported primary contributor to injury onset at the shoulder were manoeuvres characterised by loaded internal humeral rotation (33.3%), and at the hamstring were manoeuvres and postures involving front splits (100.0%). CONCLUSION: The findings indicate that pole dancers are at high risk for injuries. Future research is needed to understand the biomechani¬cal demand of manoeuvres and training characteristics of pole dancing (e.g., workload and recovery) to guide the development of preventative interventions, particularly targeted toward the shoulder and hamstring.

Immediate effects of transcutaneous electrical nerve stimulation on gait patterns in chronic stroke survivors: A single group, pretest-posttest clinical trial.

BACKGROUND: Transcutaneous electrical nerve stimulation (TENS) has been used to reduce muscle spasticity and improve locomotion in stroke survivors. We speculate that acute changes in gait performance after TENS mediate functional improvement in the long-term. However, no previous study has investigated the effect of TENS on ankle kinetics and kinematics during walking in stroke survivors. PURPOSE: We aimed to investigate whether TENS applied over the paretic leg could rapidly improve the plantar flexion moment and ankle kinematics in chronic stroke survivors with lower limb paresis. METHODS: Twenty chronic stroke survivors were recruited. They underwent 30 min of TENS over the area innervated by the common peroneal nerve on the paretic leg. Three-dimensional (3D) motion capture was performed and ankle plantar flexor spasticity was assessed before and immediately after stimulation. Ankle kinematics and kinetic and spatiotemporal data were collected using 3D motion capture. Ankle plantar flexor spasticity was assessed using the Modified Tardieu Scale. PRINCIPAL RESULTS: A significant increase in the ankle plantar flexion moment of the paretic side during the pre-swing phase was observed immediately after stimulation (p = 0.009, maximal mean difference = 0.035, 95%CI = 0.0125 to 0.0575). The step length of the paretic limb also increased significantly after stimulation (p = 0.023, mean difference = -0.02, 95%CI = -0.04 to -0.004). TENS had no immediate effect on paretic ankle spasticity, as measured by the Modified Tardieu Scale, or on other temporo-spatial parameters. CONCLUSION: The findings support the use of TENS to improve the motor function and gait pattern in chronic stroke survivors. The study indicated that the application of TENS to the paretic leg before gait training might improve rehabilitation outcomes. Future studies investigating the effects of TENS on functional outcomes, the optimal stimulation duration, and assessing spasticity using more sensitive measures are warranted.

Children with cerebral palsy have larger Achilles tendon moment arms than typically developing children.

The effectiveness of the plantarflexor muscle group to generate desired plantarflexion moments is modulated by the geometry of the Achilles tendon moment arm (ATMA). Children with cerebral palsy (CP) frequently have reduced plantarflexion function, which is commonly attributed to impaired muscle structure and function, however little attention has been paid to the potential contribution of ATMA geometry. The use of musculoskeletal modelling for the simulation of gait and understanding of gait mechanics, rely on accuracy of ATMA estimates. This study aimed to compare 3D in-vivo estimates of ATMA of adults, children with CP and typically developing (TD) children, as well as compare 3D in-vivo estimates to linearly scaled musculoskeletal model estimates. MRI scans for eight children with CP, 11 TD children and nine healthy adults were used to estimate in-vivo 3D ATMA using a validated method. A lower limb musculoskeletal model was linearly scaled to individual tibia length to provide a scaled ATMA estimate. Normalised in-vivo 3D ATMA for children with CP was 17.2% ±â€¯2.0 tibia length, which was significantly larger than for TD children (15.2% ±â€¯1.2, p = 0.013) and adults (12.5% ±â€¯0.8, p < 0.001). Scaled ATMA estimates from musculoskeletal models significantly underestimated in-vivo estimates for all groups, by up to 34.7%. The results of this study show children with CP have larger normalised 3D ATMA compared to their TD counterparts, which may have implications in understanding reduced plantarflexor function and the efficacy of surgical interventions whose aim is to modify the musculoskeletal geometry of this muscle group.

A simple but reliable method for measuring 3D Achilles tendon moment arm geometry from a single, static magnetic resonance scan.

Current methods for measuring in vivo 3D muscle-tendon moment arms generally rely on the acquisition of magnetic resonance imaging (MRI) scans at multiple joint angles. However, for patients with musculoskeletal pathologies such as fixed contractures, moving a joint through its full range of motion is not always feasible. The purpose of this research was to develop a simple, but reliable in vivo 3D Achilles tendon moment arm (ATMA) technique from a single static MRI scan. To accomplish this, for nine healthy adults (5 males, 4 females), the geometry of a cylinder was fit to the 3D form of the talus dome, which was used to estimate the talocrural flexion/extension axis, and a fifth-order polynomial fit to the line of action of the Achilles tendon. The single static scan in vivo 3D ATMA estimates were compared to estimates obtained from the same subjects at the same ankle joint angles using a previously validated 3D dynamic MRI based in vivo ATMA measurement technique. The ATMA estimates from the single scan in vivo 3D method (52.5mm±5.6) were in excellent agreement (ICC=0.912) to the validated in vivo 3D method (51.5mm±5.1). These data show reliable in vivo 3D ATMA can be obtained from a single MRI scan for healthy adult populations. The single scan, in vivo 3D ATMA technique provides researchers with a simple, but reliable method for obtaining subject-specific ATMAs for musculoskeletal modelling purposes.

Prescribing joint co-ordinates during model preparation to improve inverse kinematic estimates of elbow joint angles.

To appropriately use inverse kinematic (IK) modelling for the assessment of human motion, a musculoskeletal model must be prepared 1) to match participant segment lengths (scaling) and 2) to align the model׳s virtual markers positions with known, experimentally derived kinematic marker positions (marker registration). The purpose of this study was to investigate whether prescribing joint co-ordinates during the marker registration process (within the modelling framework OpenSim) will improve IK derived elbow kinematics during an overhead sporting task. To test this, the upper limb kinematics of eight cricket bowlers were recorded during two testing sessions, with a different tester each session. The bowling trials were IK modelled twice: once with an upper limb musculoskeletal model prepared with prescribed participant specific co-ordinates during marker registration - MRPC - and once with the same model prepared without prescribed co-ordinates - MR; and by an established direct kinematic (DK) upper limb model. Whilst both skeletal model preparations had strong inter-tester repeatability (MR: Statistical Parametric Mapping (SPM1D)=0% different; MRPC: SPM1D=0% different), when compared with DK model elbow FE waveform estimates, IK estimates using the MRPC model (RMSD=5.2±2.0°, SPM1D=68% different) were in closer agreement than the estimates from the MR model (RMSD=44.5±18.5°, SPM1D=100% different). Results show that prescribing participant specific joint co-ordinates during the marker registration phase of model preparation increases the accuracy and repeatability of IK solutions when modelling overhead sporting tasks in OpenSim.

Vector-field statistics for the analysis of time varying clinical gait data.

BACKGROUND: In clinical settings, the time varying analysis of gait data relies heavily on the experience of the individual(s) assessing these biological signals. Though three dimensional kinematics are recognised as time varying waveforms (1D), exploratory statistical analysis of these data are commonly carried out with multiple discrete or 0D dependent variables. In the absence of an a priori 0D hypothesis, clinicians are at risk of making type I and II errors in their analyis of time varying gait signatures in the event statistics are used in concert with prefered subjective clinical assesment methods. The aim of this communication was to determine if vector field waveform statistics were capable of providing quantitative corroboration to practically significant differences in time varying gait signatures as determined by two clinically trained gait experts. METHODS: The case study was a left hemiplegic Cerebral Palsy (GMFCS I) gait patient following a botulinum toxin (BoNT-A) injection to their left gastrocnemius muscle. FINDINGS: When comparing subjective clinical gait assessments between two testers, they were in agreement with each other for 61% of the joint degrees of freedom and phases of motion analysed. For tester 1 and tester 2, they were in agreement with the vector-field analysis for 78% and 53% of the kinematic variables analysed. When the subjective analyses of tester 1 and tester 2 were pooled together and then compared to the vector-field analysis, they were in agreement for 83% of the time varying kinematic variables analysed. INTERPRETATION: These outcomes demonstrate that in principle, vector-field statistics corroborates with what a team of clinical gait experts would classify as practically meaningful pre- versus post time varying kinematic differences. The potential for vector-field statistics to be used as a useful clinical tool for the objective analysis of time varying clinical gait data is established. Future research is recommended to assess the usefulness of vector-field analyses during the clinical decision making process.

Changes in muscle activation following balance and technique training and a season of Australian football.

OBJECTIVES: Determine if balance and technique training implemented adjunct to 1001 male Australian football players' training influenced the activation/strength of the muscles crossing the knee during pre-planned and unplanned sidestepping. DESIGN: Randomized Control Trial. METHODS: Each Australian football player participated in either 28 weeks of balance and technique training or 'sham' training. Twenty-eight Australian football players (balance and technique training, n=12; 'sham' training, n=16) completed biomechanical testing pre-to-post training. Peak knee moments and directed co-contraction ratios in three degrees of freedom, as well as total muscle activation were calculated during pre-planned and unplanned sidestepping. RESULTS: No significant differences in muscle activation/strength were observed between the 'sham' training and balance and technique training groups. Following a season of Australian football, knee extensor (p=0.023) and semimembranosus (p=0.006) muscle activation increased during both pre-planned sidestepping and unplanned sidestepping. Following a season of Australian football, total muscle activation was 30% lower and peak valgus knee moments 80% greater (p=0.022) during unplanned sidestepping when compared with pre-planned sidestepping. CONCLUSIONS: When implemented in a community level training environment, balance and technique training was not effective in changing the activation of the muscles crossing the knee during sidestepping. Following a season of Australian football, players are better able to support both frontal and sagittal plane knee moments. When compared to pre-planned sidestepping, Australian football players may be at increased risk of anterior cruciate ligament injury during unplanned sidestepping in the latter half of an Australian football season.

An anterior cruciate ligament injury prevention framework: incorporating the recent evidence.

Anterior cruciate ligament (ACL) injury rates have increased by ∼50% over the last 10 years. These figures suggest that ACL focused research has not been effective in reducing injury rates among community level athletes. Training protocols designed to reduce ACL injury rates have been both effective (n = 3) and ineffective (n = 7). Although a rationale for the use of exercise to reduce ACL injuries is established, the mechanisms by which they act are relatively unknown. This article provides an injury prevention framework specific to noncontact ACL injuries and the design of prophylactic training protocols. It is also apparent that feedback within this framework is needed to determine how biomechanically relevant risk factors like peak joint loading and muscular support are influenced following training. It is by identifying these links that more effective ACL injury prevention training programs can be developed, and, in turn, lead to reduced ACL injury rates in the future.

Optimizing whole-body kinematics to minimize valgus knee loading during sidestepping: implications for ACL injury risk.

The kinematic mechanisms associated with elevated externally applied valgus knee moments during non-contact sidestepping and subsequent anterior cruciate ligament (ACL) injury risk are not well understood. To address this issue, the residual reduction algorithm (RRA) in OpenSim was used to create nine subject-specific, full-body (37 degrees of freedom) torque-driven simulations of athletic males performing unplanned sidestep (UnSS) sport tasks. The RRA was used again to produce an optimized kinematic solution with reduced peak valgus knee torques during the weight acceptance phase of stance. Pre-to-post kinematic optimization, mean peak valgus knee moments were significantly reduced by 44.2 Nm (p=0.045). Nine of a possible 37 upper and lower body kinematic changes in all three planes of motion were consistently used during the RRA to decrease peak valgus knee moments. The generalized kinematic strategy used by all nine simulations to reduce peak valgus knee moments and subsequent ACL injury risk during UnSS was to redirect the whole-body center of mass medially, towards the desired direction of travel.

The NADH oxidase from Pyrococcus furiosus. Implications for the protection of anaerobic hyperthermophiles against oxidative stress.

A wealth of H(2)O-producing NADH oxidase (NOX) homologues have been discovered in the genomes of the hyperthermophilic Archaea, including two homologues in the genome of Pyrococcus furiosus which have been designated as NOX1 and NOX2. In order to investigate the function of NOX1, the structural gene encoding NOX1 was cloned from the genome of P. furiosus and expressed in Escherichia coli, and the resulting recombinant enzyme (rNOX1) was purified to homogeneity. The enzyme is a thermostable flavoprotein that can be reconstituted only with FAD. rNOX1 catalyzes the oxidation of NADH, producing both H(2)O(2) and H(2)O as reduction products of O(2) (O(2) + 1-2NADH + 1-2H(+) --> 1-2NAD(+) + H(2)O(2) or 2H(2)O). To our knowledge, this is the first NADH oxidase found to produce both H(2)O(2) and H(2)O. The enzyme exhibits a low K(m) for NADH (< 4 microm), and shows little or no reaction with NADPH. Transcriptional analyses demonstrated that NOX1 is constitutively expressed regardless of the carbon source and a single promoter was identified 25 bp upstream of the nox1 gene by primer extension. Although P. furiosus is a strict anaerobe, it may tolerate oxygen to some extent and we anticipate NOX1 to be involved in the response to oxygen at high temperatures.

The oral health of adults in the Baltimore SMSA study area.

This study, carried out in the region including Baltimore, Maryland, USA followed the pattern previously reported in the studies in Australia, New Zealand, Norway, Germany and Japan (Int. dent. J. 1976, 26, 317). In this area dental care is provided predominantly by payment of fees for service in private practice. Details are given of the oral health status in terms of caries, periodontal disease and loss of teeth for a group of 1142 adults aged between 35 and 45. Using the WHO stepwise regression analysis the significance of the previously reported predictors for each of the variables examined was determined. Finally a comparison is made between the present findings and those from an earlier National Health Survey carried out between 1960 and 1962. This suggests that while the caries situation may have improved over the past fifteen years, periodontal disease and loss of teeth have increased.

Plaque, caries, periodontal diseases, and acculturation among Yanomamö Indians, Venezuela.

The number of DM and d teeth and surfaces was recorded for 220 Yanomamö Indians from three groups of villages with different degrees of contact with Western culture. Specimens of plaque were taken from the teeth, transported in a holding solution, cultured, and examined for specific oral streptococci. In addition, the periodontal health and oral hygiene of one group of villagers were assessed using the Russell PI and the Greene & Vermillions OHIS. Caries experience among the Yanomamö was shown to be positively associated with exposure to Western culture. S. mutans was recovered with about the same frequency from specimens taken from the teeth of Indians living at all three village locations. However, the presence of S. mutans alone did not account for the disparity in dental caries scores. The examinees had abundant and persistent accumulations of soft deposits on their teeth accompanied by markedly inflamed gingival tissues. However, periodontal pockets and loss of appreciable amounts of bone did not appear as early in life nor were they as severe as reported for some other populations which practice little oral hygiene. Those disparities in the distribution of plaque-induced oral diseases between Western populations and the Yanomamö warrant further study.