Wearable IMU-Based Human Activity Recognition Algorithm for Clinical Balance Assessment Using 1D-CNN and GRU Ensemble Model

In this study, a wearable inertial measurement unit system was introduced to assess patients via the Berg balance scale (BBS), a clinical test for balance assessment. For this purpose, an automatic scoring algorithm was developed. The principal aim of this study is to improve the performance of the...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Yeon-Wook Kim, Kyung-Lim Joa, Han-Young Jeong, Sangmin Lee
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	balance assessment data augmentation gated recurrent unit human activity recognition inertial measurement unit one-dimensional convolutional neural network Chemical technology TP1-1185
Acceso en línea:	https://doaj.org/article/f9cc591b69dc431e8e3fd59e96bebaa4
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:f9cc591b69dc431e8e3fd59e96bebaa4
record_format	dspace
spelling	oai:doaj.org-article:f9cc591b69dc431e8e3fd59e96bebaa42021-11-25T18:58:03ZWearable IMU-Based Human Activity Recognition Algorithm for Clinical Balance Assessment Using 1D-CNN and GRU Ensemble Model10.3390/s212276281424-8220https://doaj.org/article/f9cc591b69dc431e8e3fd59e96bebaa42021-11-01T00:00:00Zhttps://www.mdpi.com/1424-8220/21/22/7628https://doaj.org/toc/1424-8220In this study, a wearable inertial measurement unit system was introduced to assess patients via the Berg balance scale (BBS), a clinical test for balance assessment. For this purpose, an automatic scoring algorithm was developed. The principal aim of this study is to improve the performance of the machine-learning-based method by introducing a deep-learning algorithm. A one-dimensional (1D) convolutional neural network (CNN) and a gated recurrent unit (GRU) that shows good performance in multivariate time-series data were used as model components to find the optimal ensemble model. Various structures were tested, and a stacking ensemble model with a simple meta-learner after two 1D-CNN heads and one GRU head showed the best performance. Additionally, model performance was enhanced by improving the dataset via preprocessing. The data were down sampled, an appropriate sampling rate was found, and the training and evaluation times of the model were improved. Using an augmentation process, the data imbalance problem was solved, and model accuracy was improved. The maximum accuracy of 14 BBS tasks using the model was 98.4%, which is superior to the results of previous studies.Yeon-Wook KimKyung-Lim JoaHan-Young JeongSangmin LeeMDPI AGarticlebalance assessmentdata augmentationgated recurrent unithuman activity recognitioninertial measurement unitone-dimensional convolutional neural networkChemical technologyTP1-1185ENSensors, Vol 21, Iss 7628, p 7628 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	balance assessment data augmentation gated recurrent unit human activity recognition inertial measurement unit one-dimensional convolutional neural network Chemical technology TP1-1185
spellingShingle	balance assessment data augmentation gated recurrent unit human activity recognition inertial measurement unit one-dimensional convolutional neural network Chemical technology TP1-1185 Yeon-Wook Kim Kyung-Lim Joa Han-Young Jeong Sangmin Lee Wearable IMU-Based Human Activity Recognition Algorithm for Clinical Balance Assessment Using 1D-CNN and GRU Ensemble Model
description	In this study, a wearable inertial measurement unit system was introduced to assess patients via the Berg balance scale (BBS), a clinical test for balance assessment. For this purpose, an automatic scoring algorithm was developed. The principal aim of this study is to improve the performance of the machine-learning-based method by introducing a deep-learning algorithm. A one-dimensional (1D) convolutional neural network (CNN) and a gated recurrent unit (GRU) that shows good performance in multivariate time-series data were used as model components to find the optimal ensemble model. Various structures were tested, and a stacking ensemble model with a simple meta-learner after two 1D-CNN heads and one GRU head showed the best performance. Additionally, model performance was enhanced by improving the dataset via preprocessing. The data were down sampled, an appropriate sampling rate was found, and the training and evaluation times of the model were improved. Using an augmentation process, the data imbalance problem was solved, and model accuracy was improved. The maximum accuracy of 14 BBS tasks using the model was 98.4%, which is superior to the results of previous studies.
format	article
author	Yeon-Wook Kim Kyung-Lim Joa Han-Young Jeong Sangmin Lee
author_facet	Yeon-Wook Kim Kyung-Lim Joa Han-Young Jeong Sangmin Lee
author_sort	Yeon-Wook Kim
title	Wearable IMU-Based Human Activity Recognition Algorithm for Clinical Balance Assessment Using 1D-CNN and GRU Ensemble Model
title_short	Wearable IMU-Based Human Activity Recognition Algorithm for Clinical Balance Assessment Using 1D-CNN and GRU Ensemble Model
title_full	Wearable IMU-Based Human Activity Recognition Algorithm for Clinical Balance Assessment Using 1D-CNN and GRU Ensemble Model
title_fullStr	Wearable IMU-Based Human Activity Recognition Algorithm for Clinical Balance Assessment Using 1D-CNN and GRU Ensemble Model
title_full_unstemmed	Wearable IMU-Based Human Activity Recognition Algorithm for Clinical Balance Assessment Using 1D-CNN and GRU Ensemble Model
title_sort	wearable imu-based human activity recognition algorithm for clinical balance assessment using 1d-cnn and gru ensemble model
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/f9cc591b69dc431e8e3fd59e96bebaa4
work_keys_str_mv	AT yeonwookkim wearableimubasedhumanactivityrecognitionalgorithmforclinicalbalanceassessmentusing1dcnnandgruensemblemodel AT kyunglimjoa wearableimubasedhumanactivityrecognitionalgorithmforclinicalbalanceassessmentusing1dcnnandgruensemblemodel AT hanyoungjeong wearableimubasedhumanactivityrecognitionalgorithmforclinicalbalanceassessmentusing1dcnnandgruensemblemodel AT sangminlee wearableimubasedhumanactivityrecognitionalgorithmforclinicalbalanceassessmentusing1dcnnandgruensemblemodel
_version_	1718410502320685056

Wearable IMU-Based Human Activity Recognition Algorithm for Clinical Balance Assessment Using 1D-CNN and GRU Ensemble Model

Ejemplares similares