Search

Normal ankle function provides a key contribution to everyday activities, particularly step/stair ascent and descent, where many falls occur. The rising to up-on-the-toes (UTT) 30 second test (UTT-30) is used in the clinical assessment of ankle muscle strength/function and endurance and is typically assessed by an observer counting the UTT movement completed. The aims of this study are: (i) to determine whether inertial measurement units (IMUs) provide valid assessment of the UTT-30 by comparing IMU-derived metrics with those from a force-platform (FP), and (ii) to describe how IMUs can be used to provide valid assessment of the movement dynamics/stability when performing a single UTT movement that is held for 5 s (UTT-stand). Twenty adults (26.2 ± 7.7 years) performed a UTT-30 and a UTT-stand on a force-platform with IMUs attached to each foot and the lumbar spine. We evaluate the agreement/association between IMU measures and measures determined from the FP. For UTT-30, IMU analysis of peaks in plantarflexion velocity and in FP’s centre of pressure (CoP) velocity was used to identify each repeated UTT movement and provided an objective means to discount any UTT movements that were not completed ‘fully’. UTT movements that were deemed to have not been completed ‘fully’ were those that yielded peak plantarflexion and CoP velocity values during the period of rising to up-on-the-toes that were below 1 SD of each participant’s mean peak rising velocity across their repeated UTT. The number of UTT movements detected by the IMU approach (23.5) agreed with the number determined by the FP (23.6), and each approach determined the same number of ‘fully’ completed movements (IMU, 19.9; FP, 19.7). For UTT-stand, IMU-derived movement dynamics/postural stability were moderately-to-strongly correlated with measures derived from the FP. Our findings highlight that the use of IMUs can provide valid assessment of UTT test(s).

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.