LAUSR

Single-task and cognitive-motor dual-task training (CMDT) has been extensively studied for the post-stroke function rehabilitation. However, little is known about the difference between the brain activities when these two types of tasks are being implemented. Therefore, neural correlates of single-cognitive/motor-task training (SCT/SMT) versus CMDT are investigated in this study. Specifically, a pure mathematical problem-solving task and a pure cycling task were designed for the SCT and SMT paradigms, respectively. For the CMDT, subjects should perform these two tasks simultaneously. Electroencephalography signals acquired during performing these two types of training are used to analyze the brain dynamic properties by event-related potential and event-related (de-)synchronization analyses. It can be seen from the results that compared with the SCT, theta event-related synchronization can be increased significantly by the CMDT (frontal theta: $p$ -value $= 5.44e-05 < 0.0001$ and temporal theta: $p$ -value $= 9.18e-06 < 0.0001$ ) and, meanwhile, the mu and central beta event-related desynchronization induced by the CMDT is significantly lower than that induced by the SMT (mu: p-value $= 2.28e-06 < 0.0001$ and central beta: p-value $= 4.37e-06 < 0.0001$ ). Therefore, it is proved in this study that the SMT can induce higher motor neural engagement than the CMDT and, meanwhile, compared with the SCT, higher cognitive neural activity can be induced by the CMDT, which can be explained by that the multitask coordination induces more neural activities. In addition, it is found that more attention has been paid on the relatively difficult task during the dual-task training. Therefore, the difficulty level of each task in the CMDT paradigm can be designed subject specific to implement the personalized training.