Working memory is an important foundation for advanced cognitive function. The paper combines the spatiotemporal advantages of electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS) to explore the neurovascular coupling mechanism of working memory. In the data analysis, the convolution matrix of time series of different trials in EEG data and hemodynamic response function (HRF) and the blood oxygen change matrix of fNIRS are extracted as the coupling characteristics. Then, canonical correlation analysis (CCA) is used to calculate the cross correlation between the two modal features. The results show that CCA algorithm can extract the similar change trend of related components between trials, and fNIRS activation of frontal pole region and dorsolateral prefrontal lobe are correlated with the delta, theta, and alpha rhythms of EEG data. This study reveals the mechanism of neurovascular coupling of working memory, and provides a new method for fusion of EEG data and fNIRS data.
In the field of brain-computer interfaces (BCIs) based on functional near-infrared spectroscopy (fNIRS), traditional subject-specific decoding methods suffer from the limitations of long calibration time and low cross-subject generalizability, which restricts the promotion and application of BCI systems in daily life and clinic. To address the above dilemma, this study proposes a novel deep transfer learning approach that combines the revised inception-residual network (rIRN) model and the model-based transfer learning (TL) strategy, referred to as TL-rIRN. This study performed cross-subject recognition experiments on mental arithmetic (MA) and mental singing (MS) tasks to validate the effectiveness and superiority of the TL-rIRN approach. The results show that the TL-rIRN significantly shortens the calibration time, reduces the training time of the target model and the consumption of computational resources, and dramatically enhances the cross-subject decoding performance compared to subject-specific decoding methods and other deep transfer learning methods. To sum up, this study provides a basis for the selection of cross-subject, cross-task, and real-time decoding algorithms for fNIRS-BCI systems, which has potential applications in constructing a convenient and universal BCI system.