Comparison of Neural Network Algorithms to Determine the Range of Motion Using Skeleton Models (original) (raw)
Related papers
Journal of Biomechanics, 1994
Artificial neural networks and a statistical method, linear discriminant analysis, were both applied to the recognition of temporal gait parameters associated with altered gait patterns. The duration of the double support and right and left single support phases were measured at seven speeds and three walking conditions. Data from 10 subjects were used to train neural networks, which were then tested using data from 10 other subjects. The overall performance of the networks was at least as high as that of linear discriminant analysis. The relative ease with which neural networks can be set up in a computer, and their discriminatory power, suggests that the technique has a useful role to playin gait analysis.
Human Gait Analysed by an Artificial Neural Network Model
2000
In this paper a model proposed by Sepulveda et al. (1) will be revised regarding the use of artificial neural networks to map EMG signal s and joint dynamics in the lower-limb. The original model will be used to analyse other aspects of human gait, like muscle recruitment, movement patterns and to study a problem from a patient with a
Human Gait Analysis and Classification Based on Neural Networks and Fuzzy Logic
Solid State Phenomena, 2009
Human gait analysis and classification is the process of identifying individuals by their walking manners. Computerized gait analysis using neural networks and fuzzy logic has become an integral part of the treatment decision-making process. Authors proposed the integration of kinetic data, more specifically power joints in combination with neural networks and fuzzy logic. It is a relatively new addition to other types of data including temporal and stride parameters. The performance of our approach was verified in laboratory for motion analysis. The obtained results are satisfying.
The use of neural networks to recognize patterns of human movement: gait patterns
Clinical Biomechanics, 1995
Artificial neural networks and a statistical method, linear discriminant analysis, were both applied to the recognition of temporal gait parameters associated with altered gait patterns. The duration of the double support and right and left single support phases were measured at seven speeds and three walking conditions. Data from 10 subjects were used to train neural networks, which were then tested using data from 10 other subjects. The overall performance of the networks was at least as high as that of linear discriminant analysis. The relative ease with which neural networks can be set up in a computer, and their discriminatory power, suggests that the technique has a useful role to playin gait analysis.
Journal of Physiotherapy & Physical Rehabilitation
Objective: The objective of the proposed research is to develop a methodology for modeling and evaluation of human motions, which will potentially benefit patients undertaking a physical rehabilitation therapy (e.g., following a stroke, or due to other medical conditions). The ultimate aim is to allow patients to perform home-based rehabilitation exercises using a sensory system for capturing the motions, where an algorithm will retrieve the trajectories of a patient's exercises, will perform data analysis by comparing the performed motions to a reference model of prescribed motions, and will send the analysis results to the patient's physician with recommendations for improvement. Methods: The modeling approach employs an artificial neural network, consisting of layers of recurrent neuron units and layers of neuron units for estimating a mixture density function over the spatio-temporal dependencies within the human motion sequences. Input data are sequences of motions related to a prescribed exercise by a physiotherapist to a patient, and recorded with a motion capture system. An autoencoder subnet is employed for reducing the dimensionality of captured sequences of human motions, complemented with a mixture density subnet for probabilistic modeling of the motion data using a mixture of Gaussian distributions. Results: The proposed neural network architecture produced a model for sets of human motions represented with a mixture of Gaussian density functions. The mean log-likelihood of observed sequences was employed as a performance metric in evaluating the consistency of a subject's performance relative to the reference dataset of motions. A publically available dataset of human motions captured with Microsoft Kinect was used for validation of the proposed method. Conclusion: The article presents a novel approach for modeling and evaluation of human motions with a potential application in home-based physical therapy and rehabilitation. The described approach employs the recent progress in the field of machine learning and neural networks in developing a parametric model of human motions, by exploiting the representational power of these algorithms to encode nonlinear input-output dependencies over long temporal horizons.
Procedia Manufacturing, 2015
Previously developed affordable three-dimensional (3D) Motion Analyzer System have been employed to obtain spatio-temporal gait parameters and 3D kinematics of upper body motion of both normal subjects as well as those with spinal abnormalities. However, occlusion problems have hindered the acquisition of the data. In this work, several modifications to the 3D Motion Analyzer System to improve its efficacy are proposed. First modification is the improvement in markers tracking module to overcome the occlusion problem, and the second one is automation of the subjects' anthropometry processing to minimize the possibility of error in data processing. The improved systems are then utilized to obtain 3D gait parameters and upper body motion during gait of 50 male and 50 female subjects as part of a continuous effort to establish Indonesia gait database. Prior to data acquisition, the subjects' anthropometry data and body posture are examined to ascertain normalcy. The subjects' weight and height are also assessed to ensure that they are in normal range according to Body-Mass Index (BMI) criteria. We have compared our results to those obtained in literature. The spatio-temporal and gait parameters of the subjects are in agreement with those found in literature. Furthermore, the improvements have been successfully implemented to overcome the occlusion problem and improve the program efficiency through the addition of automation of input data from a source file. Overall, the parameters obtained from this research show that the 3D Motion Analyzer system would serve the purpose of gait parameters determination well. Hence, the system has the potential for utilization as a medical diagnostic tool.
Reconstruction of gait biomechanical parameters using cyclograms and artificial neural networks
Research on Biomedical Engineering
Introduction: Historically, assessing the quality of human gait has been a difficult process. Advanced studies can be conducted using modern 3D systems. However, due to their high cost, usage of these 3D systems is still restricted to research environments. 2D systems offer simpler and more affordable solutions. Methods: In this study, the gait of 40 volunteers walking on a treadmill was recorded in the sagittal plane, using a 2D motion capture system. The extracted joint angles data were used to create cyclograms. Sections of the cyclograms were used as inputs to artificial neural networks (ANNs), since they can represent the kinematic behavior of the lower body. This allowed for prediction of future states of the moving body. Results: The results indicate that ANNs can predict the future states of the gait with high accuracy. Both single point and section predictions were successfully performed. Pearson's correlation coefficient and matched-pairs t-test ensured that the results were statistically significant. Conclusion: The combined use of ANNs and simple, accessible hardware is of great value in clinical practice. The use of cyclograms facilitates the analysis, as several gait characteristics can be easily recognized by their geometric shape. The predictive model presented in this paper facilitates generation of data that can be used in robotic locomotion therapy as a control signal or feedback element, aiding in the rehabilitation process of patients with motor dysfunction. The system proposes an interesting tool that can be explored to increase rehabilitation possibilities, providing better quality of life to patients.
2014 International Joint Conference on Neural Networks (IJCNN), 2014
The objective of this study is to investigate the use of electromyography (EMG) signals and video based soft tissue deformation (STD) analysis for identifying the gait patterns of healthy and injured subjects. The system includes a wireless surface electromyography (EMG) sensor unit and two video camera systems for measuring the neuromuscular activity of lower limb muscles, and a custom-developed artificial neural network based intelligent system software for identifying the gait patterns of subjects during walking activity. The system uses root mean square (RMS) value of EMG signals and soft tissue deformation parameter (STDP) as the input features. In order to estimate the STD during a muscular contraction while walking, flexible triangular meshes are built on reference points. The positions of these selected points are evaluated by applying the block matching motion estimation technique. Based on the extracted features, multilayer feed-forward backpropagation networks (FFBPNNs) with different network training functions were designed and their classification performances were compared. The system has been tested for a group of healthy and injured subjects. The results showed that FFBPNN with Levenberg-Marquardt training function provided better prediction behavior (98% overall accuracy) as compared to FFBPNN with other training functions for gait patterns identification based on RMS value of EMG and STDP.
Motion segmentation method for hybrid characteristic on human motion
Journal of Biomechanics, 2009
Motion segmentation and analysis are used to improve the process of classification of motion and information gathered on repetitive or periodic characteristic. The classification result is useful for ergonomic and postural safety analysis, since repetitive motion is known to be related to certain musculoskeletal disorders. Past studies mainly focused on motion segmentation on particular motion characteristic with certain prior knowledge on static or periodic property of motion, which narrowed method's applicability. This paper attempts to introduce a method to tackle human joint motion without having prior knowledge. The motion is segmented by a two-pass algorithm. Recursive least square (RLS) is firstly used to estimate possible segments on the input human-motion set. Further, period identification and extra segmentation process are applied to produce meaningful segments. Each of the result segments is modeled by a damped harmonic model, with frequency, amplitude and duration produced as parameters for ergonomic evaluation and other human factor studies such as task safety evaluation and sport analysis. Experiments show that the method can handle periodic, random and mixed characteristics on human motion, which can also be extended to the usage in repetitive motion in workflow and irregular periodic motion like sport movement.