Table 7.

Summary of fused EEG-fNIRS studies for motor task.

References	Main finding
Fazli et al. (2012)	Concurrent measurements of EEG and fNIRS can significantly improve the BCI systems classification accuracy and performance for sensory-motor rhythm.
Buccino et al. (2016)	Classification of four different hand movements is executed. Different features were compared to diminish the fNIRS delay in change detection using common spatial patterns and genetic algorithms.
Ge et al. (2017)	The study stepped forward toward real-time BCI application by using a few EEG and fNIRS channels to improve the hybrid BCI system's classification accuracy for the imaginary motor task by improving the signal acquisition (source analysis) and signal processing (phase-space reconstruction).
Li et al. (2017)	The classification accuracy for hybrid EEG-fNIRS is enhanced by integrating their complementary properties and early temporal features.
Khan M. J. et al. (2018)	A novel classifier based on a modified vector phase diagram is proposed for the finger-tapping task. The results suggest an enhancement in classification accuracy with the proposed method using a time of 1.5 s.
Chiarelli et al. (2018)	DNNs show better classification accuracy for EEG-fNIRS recording than LDA and SVM while performing left and right-hand imagery tasks.
Kwon et al. (2020)	The study proves the feasibility of achieving higher classification with less EEG electrodes and fNIRS optodes than the bulky individual EEG and fNIRS based BCI system.