“…Hence, although psychological assessment and questionnaires are probably the most accepted and validated methods in medical contexts [ 14 ], their inclusion in the automated prediction of the BCI skills remains very rare due to their disputed reliability and reproducibility [ 15 ]. For enhancing the predictive utility, the joint analysis of different imaging modalities is achieved, which may explain the discovered relationships between anatomical, functional, and electrophysiological properties of the brain [ 16 , 17 ]. Nonetheless, besides those issues that may arise by the questionary implementation, research endeavors of multimodal analysis pose a challenging problem in terms of combining categorical data with imaging measurements, facing the following restrictions [ 18 , 19 ]: Different spatial and temporal sampling rates, noninstantaneous and nonlinear coupling, low signal-to-noise ratios, a lack of interpretable results, and the optimal combination of individual modalities is still undetermined, as well as effective dimensionality reduction to enhance the discriminability of extracted multi-view features [ 20 ].…”