Transcranial Magnetic Stimulation

The electromyography (EMG) responses of suprahyoid muscle were detected by two pairs of bipolar silver-silver chloride electrodes (Yiruide, Wuhan, China), which is placed on the left and right sides of the suprahyoid muscle groups projection area. A pair of electrodes was attached to the abductor muscle of the thumb for measuring the threshold of motion. All electrodes were connected to an EMG recording system (Yiruide, Wuhan, China).

Focal TBS was delivered through a Magstim super rapid stimulator (Yiruide medical equipment Co., Wuhan, China) connected to a figure-of-eight coil 70 mm in diameter. Neuronavigation (Softaxi Optic, Canada, NDI) was used to position the stimulator at the motor cortex of the suprahyoid muscles. The subject was seated on a chair with armrest, and the coil was placed toward the target hemisphere according to navigation system and was kept in contact with the scalp closely. Single pulse stimulation is triggered from 60% of maximum output intensity, and the stimulation intensity is gradually increased until a significant left abduction activity is induced, and then the stimulation intensity is maintained, each time with a slight distance of 0.5–1.0 cm. Moving the coil for five consecutive stimulations, the position of the maximal MEP amplitude and the shortest latency period is considered to be the maximum motion stimulation zone of the left thumb abductor muscle. Fix the coil and gradually reduce the stimulation intensity until at least five times of 10 consecutive stimulations can induce MEP of ≥50 μV in the left thumb abductor muscle. This stimulation intensity is the resting motion threshold of the subject (Rest Motor Threshold, RMT). Move the coil to the anterolateral side and give single pulse stimulation with 70% output intensity. Move the coil slightly at a distance of 0.5–1.0 cm each time. The position of the MEP amplitude induced by the five consecutive stimulations, which is regarded as the best stimulation point of the suprahyoid muscle groups. The position of the MEP optimal stimulation point on the left and right suprahyoid muscle groups is preserved by the nerve positioning navigation system to ensure that the subsequent stimulation sites are consistent.

Briefly, we verified the stimulating protocol efficacy by measuring MEPs. Repetitive TMS (rTMS) was delivered using the TBS paradigm which consisted of bursts that contained three pulses at 50 Hz repeated at 5 Hz. During the iTBS, each burst of burst stimuli consisted of three consecutive pulses, 2 s stimulation, 8 s intermittent, and repeated 20 times for approximately 190 s (600 pulses). For the cTBS paradigm, each burst of burst stimuli consisted of three consecutive pulses with 200 bursts of intermittent stimulation for a total of 600 pulses with a sustained stimulation time of 40.04 s (Bertini et al., 2010).

The three groups were subjected to different protocols in the current study; in group 1 and group 2, cTBS and iTBS were positioned on the left cortex of the suprahyoid muscles. It was reported that the majority of subjects lateralized to the left hemisphere for the pharynx, right suprahyoid and left suprahyoid muscle sites. Therefore, in the present study, we chose the left hemisphere as the target. In group 3, after the cTBS placed on the left cortex of suprahyoid muscles, the iTBS was immediately delivered on the right suprahyoid muscle cortex. In this protocol, the cTBS was used to create a “virtual lesion” in the left side swallowing motor cortex and the iTBS was placed on the right side to explore whether iTBS could reverse the inhibitory effect of cTBS in the contralateral hemisphere.

The baseline and post-TBS MR studies were conducted within an interval of 2 h in the same day. Participants were scanned on a Siemens Verio 3.0 T scanner (Siemens, Erlangen, Germany). A high resolution T1-weighted images were obtained in an axial orientation [repetition time (TR) = 2,530 ms, echo time (TE) = 2.93 ms, flip angle (FA) = 7°, field of view (FOV) = 256 mm × 256 mm, slice thickness = 1.0 mm, no slice gap]. Functional images were obtained by using an echo-planar imaging sequence (33 axial slices, TR = 2,000 ms, TE = 21 ms, FA = 90°, FOV = 240 mm × 240 mm, matrix = 64 × 64, slice thickness = 4.0 mm, voxel size = 3.75 mm ×3.75 mm × 4.0 mm).

Image processing was performed using the DPARSF software package¹ (Chao-Gan and Yu-Feng, 2010). For each participant, the first 10 images of each dataset were discarded to allow for magnetization equilibrium and for the participants to adjust to the environment. All subjects had less than 2 mm maximum displacement in x, y, or z and 2° of angular motion during the whole fMRI scan. Then, the images were normalized to the standard SPM8 echo-planar imaging template, resampled with voxel size of 3 mm × 3 mm × 3 mm. The white matter signal, cerebrospinal fluid signal and Friston 24 motion parameters were removed by regression. Linear trend subtraction and temporal filtering (0.01–0.08 Hz) were carried out on the time series of each voxel to reduce the effects of low-frequency drifts and high-frequency respiratory and cardiac noise (Biswal et al., 1995).

The REST software package² was used to calculate the ReHo values and generate the ReHo maps. The details of ReHo analysis were described in previous studies (Zang et al., 2004; Wu et al., 2009). Briefly, the ReHo maps were produced by calculating the Kendall coefficient of concordance (KCC) which measures the similarity between the time series of a given voxel and those of its 26 nearest neighbors (Zang et al., 2004). Then, each individual ReHo map was divided by its own global mean KCC value within the brain mask for standardization (Wang et al., 2014). Finally, the standardized ReHo images were smoothed using a Gaussian kernel of full-width athalf-maximum 4.0 mm.

Demographics, including age and gender of all subjects, were analyzed using SPSS, version 16.0 (SPSS Inc., Chicago, IL, USA). The continuous variables of the three groups were compared using one-way analysis of variance (ANOVA), and the chi-square test was used to analyze the categorical data.

The ReHo comparisons were used the DPABI software package³ (Yan et al., 2016). We performed an inter-groups comparison of the ReHo maps of subjects between the post-TBS and baseline using paired t-tests in the three groups. We examined the normalized ReHo maps in a voxel-by-voxel manner with regression of age, gender and gray matter volume. The results were thresholded with p < 0.05 with a combined individual cluster size >85 voxels corrected using Monte Carlo simulations (see AlphaSim program in AFNI⁴ to minimize type I errors. The ANOVA test was performed for comparing the main effects of ReHo maps among the three groups with post-TBS using DPABI software package (Yan et al., 2016), and post hoc analysis was then conducted to investigate the differences of ReHo maps between paired groups (p < 0.05 with AlphaSim correction and a cluster size >85 voxels).