The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. ex. Some numerals are expressed as "XNUMX".
Copyrights notice
The original paper is in English. Non-English content has been machine-translated and may contain typographical errors or mistranslations. Copyrights notice
La rééducation et l'évaluation de la fonction motrice sont importantes pour les patients handicapés moteurs. Dans l'estimation de la longueur de foulée à l'aide d'une IMU fixée au pied, il est nécessaire de détecter le temps de l'état de mouvement, dans lequel l'accélération doit être intégrée. Dans notre étude précédente, des seuils d'accélération étaient utilisés pour déterminer la section d'intégration, il était donc nécessaire d'ajuster les valeurs seuils pour chaque sujet. Le but de cette étude était de développer une méthode d'estimation automatique de la longueur de foulée à l'aide d'un réseau de neurones artificiels (ANN). Dans cet article, un ANN à 4 couches avec des couches d'extraction de fonctionnalités entraînées par un encodeur automatique a été testé. De plus, les méthodes de recherche du minimum local d'accélération ou de sortie ANN après détection de la section d'état de mouvement par ANN ont été examinées. La méthode proposée a estimé la longueur de foulée pour les sujets sains avec une erreur de -1.88 ± 2.36 %, ce qui était presque la même que la méthode précédente basée sur un seuil (-0.97 ± 2.68 %). Les coefficients de corrélation entre la longueur de foulée estimée et la valeur de référence étaient respectivement de 0.981 et 0.976 pour les méthodes proposées et précédentes. Les plages d'erreur hors valeurs aberrantes étaient comprises entre -7.03 % et 3.23 %, entre -7.13 % et 5.09 % pour les méthodes proposées et précédentes, respectivement. La méthode proposée serait efficace parce que la plage d’erreurs était plus petite que la méthode conventionnelle et qu’aucun ajustement de seuil n’était nécessaire.
Yoshitaka NOZAKI
Tohoku University
Takashi WATANABE
Tohoku University
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copier
Yoshitaka NOZAKI, Takashi WATANABE, "Development of Artificial Neural Network Based Automatic Stride Length Estimation Method Using IMU: Validation Test with Healthy Subjects" in IEICE TRANSACTIONS on Information,
vol. E103-D, no. 9, pp. 2027-2031, September 2020, doi: 10.1587/transinf.2019EDL8227.
Abstract: Rehabilitation and evaluation of motor function are important for motor disabled patients. In stride length estimation using an IMU attached to the foot, it is necessary to detect the time of the movement state, in which acceleration should be integrated. In our previous study, acceleration thresholds were used to determine the integration section, so it was necessary to adjust the threshold values for each subject. The purpose of this study was to develop a method for estimating stride length automatically using an artificial neural network (ANN). In this paper, a 4-layer ANN with feature extraction layers trained by autoencoder was tested. In addition, the methods of searching for the local minimum of acceleration or ANN output after detecting the movement state section by ANN were examined. The proposed method estimated the stride length for healthy subjects with error of -1.88 ± 2.36%, which was almost the same as the previous threshold based method (-0.97 ± 2.68%). The correlation coefficients between the estimated stride length and the reference value were 0.981 and 0.976 for the proposed and previous methods, respectively. The error ranges excluding outliers were between -7.03% and 3.23%, between -7.13% and 5.09% for the proposed and previous methods, respectively. The proposed method would be effective because the error range was smaller than the conventional method and no threshold adjustment was required.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2019EDL8227/_p
Copier
@ARTICLE{e103-d_9_2027,
author={Yoshitaka NOZAKI, Takashi WATANABE, },
journal={IEICE TRANSACTIONS on Information},
title={Development of Artificial Neural Network Based Automatic Stride Length Estimation Method Using IMU: Validation Test with Healthy Subjects},
year={2020},
volume={E103-D},
number={9},
pages={2027-2031},
abstract={Rehabilitation and evaluation of motor function are important for motor disabled patients. In stride length estimation using an IMU attached to the foot, it is necessary to detect the time of the movement state, in which acceleration should be integrated. In our previous study, acceleration thresholds were used to determine the integration section, so it was necessary to adjust the threshold values for each subject. The purpose of this study was to develop a method for estimating stride length automatically using an artificial neural network (ANN). In this paper, a 4-layer ANN with feature extraction layers trained by autoencoder was tested. In addition, the methods of searching for the local minimum of acceleration or ANN output after detecting the movement state section by ANN were examined. The proposed method estimated the stride length for healthy subjects with error of -1.88 ± 2.36%, which was almost the same as the previous threshold based method (-0.97 ± 2.68%). The correlation coefficients between the estimated stride length and the reference value were 0.981 and 0.976 for the proposed and previous methods, respectively. The error ranges excluding outliers were between -7.03% and 3.23%, between -7.13% and 5.09% for the proposed and previous methods, respectively. The proposed method would be effective because the error range was smaller than the conventional method and no threshold adjustment was required.},
keywords={},
doi={10.1587/transinf.2019EDL8227},
ISSN={1745-1361},
month={September},}
Copier
TY - JOUR
TI - Development of Artificial Neural Network Based Automatic Stride Length Estimation Method Using IMU: Validation Test with Healthy Subjects
T2 - IEICE TRANSACTIONS on Information
SP - 2027
EP - 2031
AU - Yoshitaka NOZAKI
AU - Takashi WATANABE
PY - 2020
DO - 10.1587/transinf.2019EDL8227
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E103-D
IS - 9
JA - IEICE TRANSACTIONS on Information
Y1 - September 2020
AB - Rehabilitation and evaluation of motor function are important for motor disabled patients. In stride length estimation using an IMU attached to the foot, it is necessary to detect the time of the movement state, in which acceleration should be integrated. In our previous study, acceleration thresholds were used to determine the integration section, so it was necessary to adjust the threshold values for each subject. The purpose of this study was to develop a method for estimating stride length automatically using an artificial neural network (ANN). In this paper, a 4-layer ANN with feature extraction layers trained by autoencoder was tested. In addition, the methods of searching for the local minimum of acceleration or ANN output after detecting the movement state section by ANN were examined. The proposed method estimated the stride length for healthy subjects with error of -1.88 ± 2.36%, which was almost the same as the previous threshold based method (-0.97 ± 2.68%). The correlation coefficients between the estimated stride length and the reference value were 0.981 and 0.976 for the proposed and previous methods, respectively. The error ranges excluding outliers were between -7.03% and 3.23%, between -7.13% and 5.09% for the proposed and previous methods, respectively. The proposed method would be effective because the error range was smaller than the conventional method and no threshold adjustment was required.
ER -