The demographic characteristics will be collected, including sex, age, stroke duration, and lesion side; Penetration-aspiration (PAS) scale will be used to evaluate swallowing function. fNIRS will be used to detect changes in oxyhemoglobin (HbO) and deoxyhemoglobin (HbR) during rest and voluntary swallowing.

The demographic characteristics will be collected, including sex and age. fNIRS will be used to detect changes in oxyhemoglobin (HbO) and deoxyhemoglobin (HbR) during rest and voluntary swallowing.

fNIRS( NirScan Danyang Huichuang Medical Equipment Co. Ltd., China) will be used to detect changes in HbO and HbR during rest and voluntary swallowing. The swallowing procedure will be used consisted of three phases of manipulations: 30-s resting state periods, 30-s swallowing periods and 30-s resting state periods. A pre-collection will be conducted before the formal collection so that the subjects could understand the test process. After the pre-collection, the subjects will take a rest for 5 minutes before the formal collection.

The NirSpark package will be used to preprocess and analyze the fNIRS data. The steps will be used to preprocess the data including: deleting irrelevant time intervals and unrelated artifacts; turning the intensity of light into optical density; choosing the band-pass filter to filter the noise and interference signals (0.01-0.2 Hz); translating optical density to the oxygen concentration in the blood; and setting the initial time of the hemodynamic response function (HRF) to 30 s and the end time to 60 s (the time for a single block paradigm). To analyze the HbO2 time-series data, a generalized linear model (GLM) will be used. The GLM could calculate the degree of matching between the experimental and ideal HRF values for each task and participant. The beta value, which represents the channel's level of cortical activation, will be utilized to estimate the HRF prediction of the HbO2 signal and can be used to represent the HRF function's peak value.

The Network module of NirSpark software was used to extract the HbO2 concentration change at each time point, analyze the Pearson correlation coefficient of HbO2 concentration in each channel on the time series, and define it as the functional connection strength between channels through Fisher-Z conversion.

The Network module of NirSpark software was used to extract the HbO2 concentration change at each time point, analyze the Pearson correlation coefficient of HbO2 concentration in each channel on the time series, and define it as the functional connection strength between channels through Fisher-Z conversion.

The Network module of NirSpark software was used to extract the HbO2 concentration change at each time point, analyze the Pearson correlation coefficient of HbO2 concentration in each channel on the time series, and define it as the functional connection strength between channels through Fisher-Z conversion.