Visual guidance of the MI-BCI paradigm interface is a prerequisite for the communication of EEG signals between the user and the BCI. In order to investigate the impact of MI-BCI paradigm interfaces guided by different visual types on the performance of MI-BCI, and to investigate the main factors influencing the visual guidance of paradigm interfaces, four types of paradigm interface prototypes were first summarised as experimental objects, and experiments were conducted using EEG equipment, using co-spatial patterns (CSP) to extract EEG signal features and classification by support vector machines (SVM). The paradigm interfaces under different visual types of guidance and their influence on the performance of MI-BCI were analyzed and compared by combining the Least Significant Difference (LSD) test and subjective evaluation. The results showed that all four types of paradigm interfaces had significant effects on MI-BCI performance, and II-paradigm had the most significant effect on MI-BCI performance, while paradigm interfaces with relatively low visual complexity had clearer visual guidance and better accuracy of EEG signal classification and MI-BCI performance. On the basis of this, it was concluded that the size and color of the cue stimuli were the main factors affecting the visual guidance of the paradigm interface in the MI-BCI paradigm interface by using the hierarchical analysis (AHP) method. The findings of the study have important reference values for the design of the MI-BCI paradigm interface and the improvement of MI-BCI efficiency

motor imagery; brain-computer interface; Paradigm Interface