Swallowing, a sensorimotor behavior, helps propelling saliva or ingesting bolus from the mouth through the pharynx into the esophagus. It is a complex process comprised by a sequence of highly organized activities involving the brainstem, neural network, and innervations of muscles associated with the oral cavity, pharynx, and esophagus [1, 2]. During swallowing, a high level neuromuscular synergy driven by the skeletal (tongue) muscles and smooth muscles of the pharynx and esophagus are required in order to successfully execute the swallowing sequences [3, 4]. One of the most important physiological events that occur during the pharyngeal stage of swallowing is hyolaryngeal elevation. During the hyolaryngeal elevation, the hyoid and larynx elevate by the contraction of at least three submental muscles (mylohyoid, geniohyoid, and anterior digastric) to protect the airway and to help open the upper esophageal sphincter. Patients with disorders in submental muscles might be exposed to the risk of penetration and or aspiration of bolus materials [5]. The main function of infrahyoid muscles is to depress the hyoid bone and larynx during swallowing, so that the food bolus could be transmitted to the stomach though the esophagus [6], any impairment of these muscles may easily result in disability of food passing. Additionally, from the physiological point of view, EMG activity in the submental muscles most often initiated the swallow whereas the infrahyoid muscle activity most frequently terminated the swallow [7]. Muscle contraction time, muscle strength and muscle synergy all play significant roles in the swallowing process. As a consequence, any problem with these muscles may result in swallowing difficulty known as dysphagia. Dysphagia could lead to different clinical risks such as choking, significant weight loss, and aspiration pneumonia [8, 9], and has been reported to be a common problem of all age groups especially the elderly. Studies showed that approximately 15% of individuals of age 60 or above suffered from dysphagia [10]. Moreover, dysphagia could deteriorate the nutritional status of the patients and thus slow down the recovery rate from particular diseases. Therefore, early diagnoses of dysphagia could be of great importance for their daily life.

The evaluation of the normality of swallowing function plays an important role for the diagnoses of dysphagia. Currently, popular clinical methods for evaluating the swallowing functions in humans include videofluoroscopy [11], endoscopy esophageal manometry [12], ultrasonography [13], acoustic analysis [14, 15], magnetic resonance imaging (MRI) [16, 17], and so on. While these conventional approaches can examine swallowing functions mechanically and/or biomechanically, they would hardly provide electrophysiological information on muscular activities during swallowing. The electrophysiological information of swallowing muscles would be important for exploring the functions of deglutition and for evaluating the efficiency of clinical rehabilitation methods in dysphagia patients. In addition, these methods have some other certain limitations such as radioactive, complex operation, high cost, and invasive risk, which would restrict their clinical application for the screening and evaluation of swallowing functions. Myoelectric signal is an electrophysiological response of muscular activities. Surface electromyogram (sEMG), which is the myoelectric signal noninvasively acquired by the electrodes on skin surface has been widely used in a variety of clinical and biomedical applications [18]. Since swallowing is essentially a semiautomatic motor action of muscles, the sEMG has been also used to assess the swallowing functions [19], which should outperform radiographic, manometric, and endoscopic techniques for the study of deglutition. A number of previous investigations have been conducted to evaluate the swallowing functions by using sEMG recordings [20–26]. McKeown et al. used the independent component analysis method to analyze the sEMG signal measured from the neck region of several subjects to examine the muscle activation patterns during swallowing [23]. Vaiman and Eviatar recorded sEMG signals from a set of healthy children to establish a database for specific muscular activities during swallowing, and proposed the usage of sEMG technique for the assessment of patients with swallowing disorders [24, 25]. Perlman et al. explored the durations and temporal relationships of electromyographic activities with bipolar surface electrodes from muscles during swallowing of different boluses, and found that interarytenoid muscle showed a significant difference in duration between the saliva and 10 ml of water [26].

Note that most previous studies used sEMG recordings from one individual muscle or a pair of muscles to investigate the muscle activities during swallowing. The sEMG signals from a few EMG sensors (1–4 electrodes) can give us some electrophysiological information about the characteristics of the muscle contractions related to deglutition. However, a complete deglutition requires a complex sequence of oral and pharyngeal events for the passage of a bolus into the esophagus. The sEMG signals only from several sparse sensors might not physiologically provide enough dynamic properties of a whole deglutition procedure when one complete swallowing would involve the activities of a number of muscles and cranial nerves. Therefore, dynamically and completely exploring the electrophysiology of deglutition-related muscles should be vital for the assessment of swallowing functions and disorders such as dysphagia.

High-density sEMG (HD sEMG) is non-invasive technique to measure electrical muscle activity with a large number of closely spaced electrodes overlying an area of the skin [27, 28]. The method of HD sEMG have been utilized in a number of studies to obtain fine electrophysiological information on activities of muscles by using a two-dimension (2D) electrode array [29–34]. While the single-channel recording methods could disregard spatial distribution of the skin potential, the HD sEMG technology offers a new approach to obtain the temporal and spatial properties of the electrical muscle activities [35, 36]. The HD sEMG have been used in several different studies such as diagnosing neuromuscular disorders [30], estimating muscle conditions [32–34], and controlling a human–machine interface [36]. The feasibility of using HD sEMG dynamic topography to continuously visualize muscle activities of normal swallowing has been investigated in a pilot study conducted in our laboratory [37]. Our preliminary results suggested that sEMG recordings could provide the useful information about the electrophysiological characteristics of swallowing muscles. Note that only three subjects were used and just two cases, dry swallowing and swallowing a cup of water, were examined in the pilot study.

In the current study, the performance of using HD sEMG technique to physiologically evaluate swallowing functions would be systemically investigated when a number of subjects were swallowing different kinds of eatable materials. Since understanding the functions of a normal deglutition should be an essential prerequisite for early diagnosis and proper therapy of dysphagia, this study would concentrate on the evaluation of normal swallowing functions by using dynamic HD sEMG. sEMG signals generated by muscle activities during swallowing were continuously recorded with a HD two-dimension (2D) electrode array. The sEMG recordings from an entire swallowing process were properly segmented into a series of sEMG analysis windows and then the sEMG energy of each analysis window was computed to generate a 2D map. These sEMG energy maps could visually demonstrate the muscle activities in deglutition since they will present the electrophysiological responses of voluntary and involuntary muscle contractions. Some quantization parameters were calculated from each of the sEMG energy maps and used to objectively and quantitatively examine the cooperation of different muscles during the swallowing. The HD sEMG approach proposed by this study may be useful for the continuous visualization of muscular activities and accurate quantification of muscle energy distribution related to normal swallowing processes, which might provide a potential tool for screening swallowing disorders to assist precise treatment of dysphagia patients.

The primary objective of this work was to investigate the feasibility of HD sEMG maps as a new approach for the continuous visualization of swallowing process. This objective has been reached by using a 2D electrode array on neck muscles to record the sEMG during swallowing. In the current study, the HD sEMG signals were recorded with an array of closely-spaced electrodes, and were processed in the space dimension to assess global muscular activities related to swallowing. The RMS value of each channel over a time duration was calculated and was considered as a pixel located at the definite position of the electrode array. Thus a series of 2D temporal sEMG maps were constructed by the RMS values of 96 channels. The dynamic HD sEMG maps would provide both spatial and temporal properties of the electrical muscle activity and presented the muscle contraction coordination that would be closely related to the swallowing function. The distribution of sEMG intensities were visually inspected on the maps where the intensity values were represented by color tones (cold to hot color). Thus the HD sEMG maps would provide a quantification of the temporal and spatial properties of the electrical muscle activity. This technique could overcome some limitations of the previous studies that were only based on the sEMG signals recorded with a few electrodes located in an individual muscle or a muscle pair.

In this paper, the normal swallowing process has been systematically investigated by the HD sEMG signals with an attempt to establish a dynamic muscle activity pattern of normal swallowing, which might be served as a reference for screening and assessment of swallowing disorders or dysphagia in the future studies and possible clinical applications. Our results from a number of normal subjects showed that the dynamic swallowing topographic maps could present the changing process of sEMG spatial distribution during swallowing, providing a new method to analyze the muscle contraction coordination during swallowing. The sEMG potential maps mainly reflected the activities of two muscle groups: submental and infrahyoid muscles, which corresponded to the high intensity on the top and in the center of maps, respectively. Our results were consistent with those of previous studies. A swallowing process could be characterized by submental and infrahyoid muscle bursts, initiated by increased submental muscle activity and terminated by decreased infrahyoid muscle activity.

During a normal swallowing, the bolus would be transported from the oral cavity though the pharynx to the esophagus. Accordingly, the temporal sEMG energy concentration in the HD maps was in a consistent shift with the movement of the bolus. At the beginning of the swallowing task, the water was hold in the oral cavity and sitting on the dorsum of the tongue, thus the submental muscles became active and a high intensity region appeared at the top of the map. When the water was propelled from the oral cavity into the oropharynx, the energy concentration shifted downward to the region of infrahyoid muscles, accomplished by evident increase of activity amplitude. Later the larynx, together with the nasopharynx, opened accordingly resulting in a maximum in the energy map. Finally, when the water was passed down to the esophagus, the muscle activities started to decrease until a resting state was reached. The dynamic HD sEMG maps could provide spatial and temporal properties of the muscle electrical activity, and visualize the muscle contraction that closely related to the swallowing function. The HD sEMG maps could be served as a new approach for the continuous visualization of the swallowing process, and could be possibly used as a potential tool for fast screening and objective assessment of swallowing disorders or dysphagia.

The entire swallowing process can be divided into five sub-phases in normal physiological process of swallowing ease detailed analysis of swallowing [38]. In this study, five HD sEMG energy maps averaged within the time interval of each phase were constructed to show the swallowing phases corresponding to the physiological process of swallowing (Fig. 5). In phase 1, the subjects were instructed to swallow the bolus in their mouth, so the tip of the tongue moved upward and the bolus was propelled posteriorly toward the pharynx. As a result, the activities of the infrahyoid muscles decreased and those of the submental muscles increased in response to the lifting of the bolus, indicated by a high intensity region on the top of the energy map (frame #1 in Fig. 5c). In phase 2, the bolus was pushed into the oropharynx while the area of soft palate and larynx began to elevate. Consequently, the submental muscle activities got stronger, and the region with outstanding muscle activities started to expand and moved down to the center (frame #2 in Fig. 5c). In phase 3, the hyoid bone and larynx were pulled upward, and the bolus was squeezed downward through the oropharynx and further into the hypopharynx by the sequential contraction of pharyngeal constrictor muscles. So muscle activities became more intense and concentrated in the center of the map (frame #3 in Fig. 5c). Meanwhile, the energy concentration split into two symmetric areas on the left and right sides of the larynx, meaning that the energy of the swallowing activities had a nearly equivalent distribution on both sides. In order to quantitatively analyze the symmetry properties of both sides, two metrics (LRER and LRED) were calculated and the results showed strong consistency for all subjects and different swallowing tasks (with the head in the middle position). These results revealed that the muscle associated with swallowing located at the left area and the right area of the neck were activated simultaneously with nearly equivalent intensity and durability, thus producing a balanced oral pressure to aid the movement of the pharynx and larynx in normal swallowing process. This symmetry analyses were consistent with our previous study [40]. Finally, in phase 4 and 5, the cricopharyngeal muscle and the upper esophageal sphincter relaxed, as the bolus passed into the esophagus. Thus, the overall intensity of the energy maps gradually decreased and down to the resting-state level. The HD sEMG energy maps were well associated with the physiological and biomechanical principles of swallowing, making the HD sEMG method a simple and non-invasive tool to accurately differentiate phases of the swallowing. It could also help researchers and therapist better understand the dynamic muscular activities associated with each phase of normal swallowing process. Additionally, the HD sEMG energy maps was used to examine the effects of different volumes, viscosities, and head postures on normal swallowing process. The energy maps obtained from swallowing water with different volumes showed that the duration of the dry swallowing was the longest across the four volume tasks. This might be attributed to the lubrication effects of water that makes swallowing easier, thus the water swallowing took shorter time than try swallowing. Moreover, the muscular activities slightly increased when the volume of the water increased. These findings were in accordance with results from the previous studies [41–43].

The energy maps obtained from swallowing bolus of different viscosities (water, thin paste, and thick paste) showed that the time it took to complete the swallowing significantly prolonged when the viscosity increased (Fig. 6). This prolongation was usually accompanied by the increase in the number of swallowing attempts as observed in Fig. 6b, c. It suggests that multiple of retries would appear if the subjects found the bolus difficult to swallow. For the swallowing of bolus with increased viscosity, evident increase in the amplitude of muscular activities were also observed (Fig. 6). This might be attribute to larger oral pressure and stronger laryngeal elevation when swallowing bolus with higher viscosity. Similar findings were also reported in a number of previous studies that evaluated swallowing processes by using other techniques [7, 44–47], and could provide much more details because of the usage of the HD sEMG. The results supported that increasing bolus viscosity required larger motor neuron pools and resulting more muscular activities. The HD sEMG energy maps of different volumes and viscosities implied that clinical practice should consider about the influence of bolus material and volume to manage the diet of dysphagia for reducing secondary physiological impairment.

The energy map in Fig. 7 revealed the pattern changes of the swallowing process when the head was kept at different postures. When the subject turned his head to the left side, the bolus pathway shifted to the right side so the muscles of the right side became more active, indicated by more energy concentration on the right side in the HD maps. Similarly, the major muscular activities gathered at the left side when the head was turned to the right side. These findings agreed with previous study [40, 48] and could provide a potential method to improve swallowing in patients with unilateral dysphagia.

The purpose of present study was to evaluate the ability of dynamic sEMG topography to represent the process of normal swallowing. In the further study, the patients with different types of swallowing disorders will be recruited, and the muscle activity patterns of normal swallowing obtained in this study might be a reference for further investigation focused on the development of clinical application for the screening and/or evaluation of swallowing function or dysphagia.

In this study, the method of HD sEMG energy maps was proposed to visualize the continuous dynamic process when the subjects were swallowing bolus of different volume and viscosity. The results showed that a dry swallowing process seemed to take a longer time than water swallowing. The maximum sEMG activities increased significantly from swallowing liquid to thin paste and from thin pastes to thick pastes for both submental and infrahyoid areas. This study suggested that the muscular activities were nearly symmetrical between left and right sides for different swallowing tasks when the subject’s head was kept in the middle position. The energy maps of different head postures revealed that the concentration of the muscular activities shifted to the other side in oppose to the head position. The proposed method might provide a useful tool to evaluate the normality of the muscular functions, and help to pin out the possible cause of swallowing difficulties for patients with swallowing disorders.