Numerous methods for studying the prevention of falls and age-related sensorimotor degradation have been proposed and tested. Some approaches are too impractical to use with seniors or too expensive for practitioners. Practitioners desire a simple, reliable technique. The goals of this research were to develop such an approach and to apply it in exploring the effect of Tai Chi on age-related sensorimotor degradation. The method employed artificial-neural-network (ANN) models trained by using individuals’ center-of-pressure (COP) measurements and age. Ninety-six White and Chinese adults without Tai Chi training were tested. In contrast, a third group, Chinese seniors with Tai Chi training, was tested to ascertain any influence from Tai Chi on sensorimotor aging. This study supported ANN technology with COP data as a feasible tool in the exploration of sensorimotor degradation and demonstrated that Tai Chi slowed down the effects of sensorimotor aging.
Gongbing Shan, Dayna Daniels and Rongri Gu
Kristof Kipp, John Krzyszkowski and Daniel Kant-Hull
, researchers have used artificial neural networks (ANNs) for the same purposes. 8 – 10 For example, ANNs were used to successfully predict swimming performance from 4 weeks of training load data, which included weekly training volume for swim-related activities, resistance exercise, and dryland training. 8
Jianning Wu and Jue Wang
In this technical note, we investigate a combination PCA with SVM to classify gait pattern based on kinetic data. The gait data of 30 young and 30 elderly participants were recorded using a strain gauge force platform during normal walking. The gait features were first extracted from the recorded vertical directional foot– ground reaction forces curve using PCA, and then these extracted features were adopted to develop the SVM gait classifier. The test results indicated that the performance of PCA-based SVM was on average 90% to recognize young– elderly gait patterns, resulting in a markedly improved performance over an artificial neural network–based classifier. The classification ability of the SVM with polynomial and radial basis function kernels was superior to that of the SVM with linear kernel. These results suggest that the proposed technique could provide an effective tool for gait classification in future clinical applications.
Arne Jaspers, Tim Op De Beéck, Michel S. Brink, Wouter G.P. Frencken, Filip Staes, Jesse J. Davis and Werner F. Helsen
Australian football (AFL) found that artificial neural networks (ANNs), a machine learning approach, more accurately predicted the RPE in response to ELIs compared with traditional statistics. 9 Other machine learning techniques could be used for this task as well, and each technique has strengths and
Berit Steenbock, Marvin N. Wright, Norman Wirsik and Mirko Brandes
artificial neural network model. Linear regression models (LM) and linear mixed models (MLM) were created separately for the outcome variables absolute EE (absEE), relative EE (relEE), and METs and included the above mentioned 30 summary statistics as independent variables. The MLM accounts for repeated
Alexander H.K. Montoye, Kimberly A. Clevenger, Kelly A. Mackintosh, Melitta A. McNarry and Karin A. Pfeiffer
combination thereof) yield higher EE prediction accuracy, and 3) compare the accuracy of these machine learning models to three count-based EE prediction regression equations. Methods In the present study, we describe the development (calibration) of six artificial neural networks (ANNs) and then focus on the
Alexandra Valencia-Peris, José Devís-Devís, Xavier García-Massó, Jorge Lizandra, Esther Pérez-Gimeno and Carmen Peiró-Velert
Previous research shows contradictory findings on potential competing effects between sedentary screen media usage (SMU) and physical activity (PA). This study examined these effects on adolescent girls via self-organizing maps analysis focusing on 3 target profiles.
A sample of 1,516 girls aged 12 to 18 years self-reported daily time engagement in PA (moderate and vigorous intensity) and in screen media activities (TV/video/DVD, computer, and videogames), separately and combined.
Topological interrelationships from the 13 emerging maps indicated a moderate competing effect between physically active and sedentary SMU patterns. Higher SES and overweight status were linked to either active or inactive behaviors. Three target clusters were explored in more detail. Cluster 1, named temperate-media actives, showed capabilities of being active while engaging in a moderate level of SMU (TV/video/DVD mainly). In Cluster 2, named prudent-media inactives, and Cluster 3, compulsive-media inactives, a competing effect between SMU and PA emerged, being sedentary SMU behaviors responsible for a low involvement in active pursuits.
SMU and PA emerge as both related and independent behaviors in girls, resulting in a moderate competing effect. Findings support the case for recommending the timing of PA and SMU for recreational purposes considering different profiles, sociodemographic factors and types of SMU.
Joan E. Deffeyes, Regina T. Harbourne, Wayne A. Stuberg and Nicholas Stergiou
Sitting is one of the first developmental milestones that an infant achieves. Thus measurements of sitting posture present an opportunity to assess sensorimotor development at a young age. Sitting postural sway data were collected using a force plate, and the data were used to train a neural network controller of a model of sitting posture. The trained networks were then probed for sensitivity to position, velocity, and acceleration information at various time delays. Infants with typical development developed a higher reliance on velocity information in control in the anterior-posterior axis, and used more types of information in control in the medial-lateral axis. Infants with delayed development, where the developmental delay was due to cerebral palsy for most of the infants in the study, did not develop this reliance on velocity information, and had less reliance on short latency control mechanisms compared with infants with typical development.
Raviraj Nataraj, Musa L. Audu, Robert F. Kirsch and Ronald J. Triolo
This pilot study investigated the potential of using trunk acceleration feedback control of center of pressure (COP) against postural disturbances with a standing neuroprosthesis following paralysis. Artificial neural networks (ANNs) were trained to use three-dimensional trunk acceleration as input to predict changes in COP for able-bodied subjects undergoing perturbations during bipedal stance. Correlation coefficients between ANN predictions and actual COP ranged from 0.67 to 0.77. An ANN trained across all subject-normalized data was used to drive feedback control of ankle muscle excitation levels for a computer model representing a standing neuroprosthesis user. Feedback control reduced average upper-body loading during perturbation onset and recovery by 42% and peak loading fby 29% compared with optimal, constant excitation.
Jonathan D. Bartlett, Fergus O’Connor, Nathan Pitchford, Lorena Torres-Ronda and Samuel J. Robertson
The aim of this study was to quantify and predict relationships between rating of perceived exertion (RPE) and GPS training-load (TL) variables in professional Australian football (AF) players using group and individualized modeling approaches.
TL data (GPS and RPE) for 41 professional AF players were obtained over a period of 27 wk. A total of 2711 training observations were analyzed with a total of 66 ± 13 sessions/player (range 39–89). Separate generalized estimating equations (GEEs) and artificial-neural-network analyses (ANNs) were conducted to determine the ability to predict RPE from TL variables (ie, session distance, high-speed running [HSR], HSR %, m/min) on a group and individual basis.
Prediction error for the individualized ANN (root-mean-square error [RMSE] 1.24 ± 0.41) was lower than the group ANN (RMSE 1.42 ± 0.44), individualized GEE (RMSE 1.58 ± 0.41), and group GEE (RMSE 1.85 ± 0.49). Both the GEE and ANN models determined session distance as the most important predictor of RPE. Furthermore, importance plots generated from the ANN revealed session distance as most predictive of RPE in 36 of the 41 players, whereas HSR was predictive of RPE in just 3 players and m/min was predictive of RPE in just 2 players.
This study demonstrates that machine learning approaches may outperform more traditional methodologies with respect to predicting athlete responses to TL. These approaches enable further individualization of load monitoring, leading to more accurate training prescription and evaluation.