Exploiting pretrained CNN models for the development of an EEG-based robust BCI framework. (April 2022)

• Record Type: Journal Article
• Authors: Sadiq, Muhammad Tariq; Aziz, Muhammad Zulkifal; Almogren, Ahmad; Yousaf, Adnan; Siuly, Siuly; Rehman, Ateeq Ur
• Is Part Of: Computers in biology and medicine. Volume 143(2022)
• Publication Date: 2022-04
• Subjects: Brain-computer interface -- Convolution neural network -- Transfer learning -- Optimizers -- Learning rates -- Classification; Medicine -- Data processing -- Periodicals; Biology -- Data processing -- Periodicals; 610.285
• DOI: 10.1016/j.compbiomed.2022.105242
• Languages: English
• ISSNs: 0010-4825

Abstract

Abstract: Identifying motor and mental imagery electroencephalography (EEG) signals is imperative to realizing automated, robust brain-computer interface (BCI) systems. In the present study, we proposed a pretrained convolutional neural network (CNN)-based new automated framework feasible for robust BCI systems with small and ample samples of motor and mental imagery EEG training data. The framework is explored by investigating the implications of different limiting factors, such as learning rates and optimizers, processed versus unprocessed scalograms, and features derived from untuned pretrained models in small, medium, and large pretrained CNN models. The experiments were performed on three public datasets obtained from BCI Competition III. The datasets were denoised with multiscale principal component analysis, and time-frequency scalograms were obtained by employing a continuous wavelet transform. The scalograms were fed into several variants of ten pretrained models for feature extraction and identification of different EEG tasks. The experimental results showed that ShuffleNet yielded the maximum average classification accuracy of 99.52% using an RMSProp optimizer with a learning rate of 0.000 1. It was observed that low learning rates converge to more optimal performances compared to high learning rates. Moreover, noisy scalograms and features extracted from untuned networks resulted in slightly lower performance than denoised scalograms and tuned networks, respectively. The overall results suggest that pretrained models are robust when identifying EEG signals because of their ability to preserve the time-frequency structure of EEG signals and promising classification outcomes. Highlights: We exploit ten pre-trained CNN models for EEG-based diverse BCI applications. Experiments were conducted to explore noises, learning rates, and optimizes effects. Features from untuned pre-trained CNN models were tested with benchmark classifiers. The Shufflenet yields the highest classification accuracy of 99.52% using an rmsprop optimizer and 0.0001 LR. The fine-tuned classification strategy prevailed over the untuned features extraction and classification approach.