Abstract
Inter-subject transfer learning is a long-standing problem in brain-computer interfaces (BCIs) and has not yet been fully realized due to high inter-subject variability in the brain signals related to motor imagery (MI). The recent success of deep learning-based algorithms in classifying different brain signals warrants further exploration to determine whether it is feasible for the inter-subject continuous decoding of MI signals to provide contingent neurofeedback which is important for neurorehabilitative BCI designs. In this paper, we have shown how a convolutional neural network (CNN) based deep learning framework can be used for inter-subject continuous decoding of MI related electroencephalographic (EEG) signals using the novel concept of Mega Blocks for adapting the network against inter-subject variabilities. These Mega Blocks have the capacity to repeat a specific architectural block several times such as one or more convolutional layers in a single Mega Block. The parameters of such Mega Blocks can be optimized using Bayesian hyperparameter optimization. The results, obtained on the publicly available BCI competition IV-2b dataset, yields an average inter-subject continuous decoding accuracy of 71.49% (κ = 0.42) and 70.84% (κ = 0.42) for two different training methods such as adaptive moment estimation (Adam) and stochastic gradient descent (SGDM), respectively, in 7 out of 9 subjects. Our results show for the first time that it is feasible to use CNN based architectures for inter-subject continuous decoding with a sufficient level of accuracy for developing calibration-free MI-BCIs for practical purposes.
Highlights
The practical applications of brain-computer interfaces are often hindered by the need for repeated calibration for each individual participant due to large inter-subject variability in the EEG signal
The combination of a 2 s time window and 200 ms shift makes 11 segments within the 5 s motor imagery (MI) period producing 11 images in a single trial. This design would be useful when it comes to providing continuous neurofeedback in a more intuitive way and for which having low latency is an essential criterion (Foldes et al, 2015). Unlike these previous studies (Foldes et al, 2015; Chowdhury et al, 2018a) which are primarily based on within-subject learning, we have shown how continuous feedback could be incorporated into a convolutional neural network (CNN) based inter-subject transfer learning setting which can contribute to calibrationfree neurorehabilitative brain-computer interfaces (BCIs) designs without compromising the richness of the neurofeedback
The performance of stochastic gradient descent method (SGDM) for this category resulted in an average classification accuracy of 73.13% ± 14.82, while the maximum accuracy was observed for participant 4 (91.14%) and the minimum observed for participant 3 (54.61%)
Summary
The practical applications of brain-computer interfaces are often hindered by the need for repeated calibration for each individual participant due to large inter-subject variability in the EEG signal. With regards to neurorehabilitation especially, the time-consuming calibration process leads to user frustration and a lack of motivation which can hinder the recovery process. This is evident from the work of Morone and colleagues who found a significant correlation between motivation and BCI performance (Morone et al, 2015) which is further found to be strongly correlated with motor recovery (Bundy et al, 2017). Recent studies utilized some BCI performance Predictors to augment the transfer learning process (Saha et al, 2018; Saha et al, 2019)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
