TY - JOUR
T1 - Improving inertial sensor-based activity recognition in neuro-logical populations
AU - Celik, Yunus
AU - Aslan, M. Fatih
AU - Sabanci, Kadir
AU - Stuart, Samuel
AU - Woo, Wai Lok
AU - Godfrey, Alan
N1 - Funding information: Yunus Celik receives Ph.D. studentship support from the Turkish Ministry of National Education. Dr. Stuart is supported, in part, by the Parkinson’s Foundation (PF-FBS-1898, PF-CRA2073).
PY - 2022/12/15
Y1 - 2022/12/15
N2 - Inertial sensor-based human activity recognition (HAR) has a range of healthcare applications as it can indicate the overall health status or functional capabilities of people with impaired mobility. Typically, artificial intelligence models achieve high recognition accuracies when trained with rich and diverse inertial datasets. However, obtaining such datasets may not be feasible in neurological populations due to, e.g., impaired patient mobility to perform many daily activities. This study proposes a novel framework to overcome the challenge of creating rich and diverse datasets for HAR in neurological populations. The framework produces images from numerical inertial time-series data (initial state) and then artificially augments the number of produced images (enhanced state) to achieve a larger dataset. Here, we used convolutional neural network (CNN) architectures by utilizing image input. In addition, CNN enables transfer learning which enables limited datasets to benefit from models that are trained with big data. Initially, two benchmarked public datasets were used to verify the framework. Afterward, the approach was tested in limited local datasets of healthy subjects (HS), Parkinson’s disease (PD) population, and stroke survivors (SS) to further investigate validity. The experimental results show that when data augmentation is applied, recognition accuracies have been increased in HS, SS, and PD by 25.6%, 21.4%, and 5.8%, respectively, compared to the no data augmentation state. In addition, data augmentation contributes to better detection of stair ascent and stair descent by 39.1% and 18.0%, respectively, in limited local datasets. Findings also suggest that CNN architectures that have a small number of deep layers can achieve high accuracy. The implication of this study has the potential to reduce the burden on participants and researchers where limited datasets are accrued.
AB - Inertial sensor-based human activity recognition (HAR) has a range of healthcare applications as it can indicate the overall health status or functional capabilities of people with impaired mobility. Typically, artificial intelligence models achieve high recognition accuracies when trained with rich and diverse inertial datasets. However, obtaining such datasets may not be feasible in neurological populations due to, e.g., impaired patient mobility to perform many daily activities. This study proposes a novel framework to overcome the challenge of creating rich and diverse datasets for HAR in neurological populations. The framework produces images from numerical inertial time-series data (initial state) and then artificially augments the number of produced images (enhanced state) to achieve a larger dataset. Here, we used convolutional neural network (CNN) architectures by utilizing image input. In addition, CNN enables transfer learning which enables limited datasets to benefit from models that are trained with big data. Initially, two benchmarked public datasets were used to verify the framework. Afterward, the approach was tested in limited local datasets of healthy subjects (HS), Parkinson’s disease (PD) population, and stroke survivors (SS) to further investigate validity. The experimental results show that when data augmentation is applied, recognition accuracies have been increased in HS, SS, and PD by 25.6%, 21.4%, and 5.8%, respectively, compared to the no data augmentation state. In addition, data augmentation contributes to better detection of stair ascent and stair descent by 39.1% and 18.0%, respectively, in limited local datasets. Findings also suggest that CNN architectures that have a small number of deep layers can achieve high accuracy. The implication of this study has the potential to reduce the burden on participants and researchers where limited datasets are accrued.
KW - human activity recognition
KW - inertial measurement units
KW - data augmentation
KW - convolutional neural networks
UR - http://www.scopus.com/inward/record.url?scp=85144553606&partnerID=8YFLogxK
U2 - 10.3390/s22249891
DO - 10.3390/s22249891
M3 - Article
C2 - 36560259
AN - SCOPUS:85144553606
SN - 1424-3210
VL - 22
JO - Sensors
JF - Sensors
IS - 24
M1 - 9891
ER -