ABSTRACT
Brain Computer-Interface (BCI) technology plays a considerable role in the control of rehabilitation or peripheral devices for stroke patients. This is particularly due to their inability to control such devices from their inherent physical limitations after such an attack. More often than not, the control of such devices exploits electroencephalogram (EEG) signals. Nonetheless, it is worth noting that the extraction of the features and the classification of the signals is non-trivial for a successful BCI system. The use of Transfer Learning (TL) has been demonstrated to be a powerful tool in the extraction of essential features. However, the employment of such a method towards BCI applications, particularly in regard to EEG signals, are somewhat limited. The present study aims to evaluate the effectiveness of different TL models in extracting features for the classification of wink-based EEG signals. The extracted features are classified by means of fine-tuned Random Forest (RF) classifier. The raw EEG signals are transformed into a scalogram image via Continuous Wavelet Transform (CWT) before it was fed into the TL models, namely InceptionV3, Inception ResNetV2, Xception and MobileNet. The dataset was divided into training, validation, and test datasets, respectively, via a stratified ratio of 60:20:20. The hyperparameters of the RF models were optimised through the grid search approach, in which the five-fold cross-validation technique was adopted. The optimised RF classifier performance was compared with the conventional TL-based CNN classifier performance. It was demonstrated from the study that the best TL model identified is the Inception ResNetV2 along with an optimised RF pipeline, as it was able to yield a classification accuracy of 100% on both the training and validation dataset. Therefore, it could be established from the study that a comparable classification efficacy is attainable via the Inception ResNetV2 with an optimised RF pipeline. It is envisaged that the implementation of the proposed architecture to a BCI system would potentially facilitate post-stroke patients to lead a better life quality.
ABSTRACT
Minimal information is known about the three-dimensional (3D) ground reaction forces (GRF) on the gait patterns of individuals with autism spectrum disorders (ASD). The purpose of this study was to investigate whether the 3D GRF components differ significantly between children with ASD and the peer controls. 15 children with ASD and 25 typically developing (TD) children had participated in the study. Two force plates were used to measure the 3D GRF data during walking. Time-series parameterization techniques were employed to extract 17 discrete features from the 3D GRF waveforms. By using independent t-test and Mann-Whitney U test, significant differences (p<0.05) between the ASD and TD groups were found for four GRF features. Children with ASD demonstrated higher maximum braking force, lower relative time to maximum braking force, and lower relative time to zero force during mid-stance. Children with ASD were also found to have reduced the second peak of vertical GRF in the terminal stance. These major findings suggest that children with ASD experience significant difficulties in supporting their body weight and endure gait instability during the stance phase. The findings of this research are useful to both clinicians and parents who wish to provide these children with appropriate treatments and rehabilitation programs.
