**2. Material and method**

We aimed to elucidate the relationship between changes in the electrical threshold of sensa‐ tions associated with various orofacial symptoms and the recovery processes of areas affected by orofacial disorders depending on the oral rehabilitation technique. Thus, we investigated the relationship between modified VF assessments and sensory electrical threshold evoked on the soft palate. In particular, electrical stimuli are appealed by a loose press feeling in the stimulus area of subjects, not painful sensation.

#### **2.1. Subjects**

**First experiment:** We examined 11 patients (8 males, 3 females; age range, 30-66 years) with various disorders in orofacial regions (facial muscles, n=4; lingual muscles, n=5; hypoesthesia ofthe soft palate, n=2)(Table 1).All medicaltreatments were performed over a long time period (range, 10-80 months). We compared the relationship between the degree of recovery follow‐ ing oralrehabilitation and the electricalthreshold of sensation in the deficient orofacialregions.

The various disorders among the 11 patients were as follows: acoustic nerve tumor (n=1), diabetes (n=1), cerebral infarct (n=4), neck tumor (n=4), and facial nerve palsy (n=1) (Table 1). Guidance regarding an accepted way of stretch training fitting various disorder pats, such as gum rubbing, the Mendelsohn maneuver, thermal tactile stimulation, the head lift exercise (Shaker exercise), and the tongue holding maneuver were provided. Training involved procedures suitable for each patient after obtaining first-person informed consent. Patients performed the training method best suited to their disease twice daily. The thermal tactile stimulation was performed when the patients visited the hospital, and the relationships between the electrical threshold of sensation from one month to the next as well as the degree of recovery according to the training condition, including hearing investigation were examined. medical treatments were performed over a long time period (range, 10-80 months). We compared the relationship between the degree of recovery following oral rehabilitation and the electrical threshold of sensation in the deficient orofacial regions. The various disorders among the 11 patients were as follows: acoustic nerve tumor (n=1), diabetes (n=1), cerebral infarct (n=4), neck tumor (n=4), and facial nerve palsy (n=1) (Table 1). Guidance regarding an accepted way of stretch training fitting various disorder pats, such as gum rubbing, the Mendelsohn maneuver, thermal tactile stimulation, the head lift exercise (Shaker exercise), and the tongue holding maneuver were provided. Training involved procedures suitable for each patient after obtaining first-person informed consent. Patients performed the training method best suited to their disease twice daily. The thermal tactile stimulation was performed when the patients visited the hospital, and the relationships between the electrical threshold of sensation from one month to the next as well as the

degree of recovery according to the training condition, including hearing investigation were examined.

disorders in orofacial regions (facial muscles, n=4; lingual muscles, n=5; hypoesthesia of the soft palate, n=2) (Table 1). All

Finally, we examined the relationship between the total scores for all items in the oral and pharyngeal phases and the threshold of electrical sensation on the soft palate. We hypothesized that if electrical threshold measurement can be substituted for VF assessments, radiation exposure to patients can be reduced. Furthermore, electrical threshold stimuli on the soft palate and other areas are produced by an appeal of a loose press feeling in the stimulus area of subjects, not

We aimed to elucidate the relationship between changes in the electrical threshold of sensations associated with various orofacial symptoms and the recovery processes of areas affected by orofacial disorders depending on the oral rehabilitation technique. Thus, we investigated the relationship between modified VF assessments and sensory electrical threshold evoked on the soft palate. In particular, electrical stimuli are appealed by a loose press feeling in the stimulus

painful sensation.

**2.1. Subjects**

**2. Material and method**

area of subjects, not painful sensation.


Table 1. Patient characteristics in the first experiment. F: female, M: male. **Table 1.** Patient characteristics in the first experiment. F: female, M: male.

Many researchers have evaluated VF results based on the Videofluorographic Examination of Swallowing Worksheet developed by Logemann [8]. However, because the inspection items on this worksheet comprise many measurement items, numerous hospitals conduct their own modified VF assessments [2, 3, 5-7, 14]. Based on experiences in other hospitals, 11 key events were identified for the assessment of VF: bolus formation, tongue-to-palate contact, premature bolus loss, residue in the oral cavity, and oral transit time in the oral phase; and lift in the soft palate, triggering of the pharyngeal swallow, vallecula residue, pyriform sinus residue, pharyngeal transit time, and aspiration in the pharyngeal phase. Furthermore, these items were classified into three levels: grade 0 as normal, grade 1 as inadequate, and grade 2 as a true abnormality. The total scores for all items in the oral and pharyngeal phases separately are calculated and a higher sum in each phase represented a more serious condition is presumed. Electrical thresholds of sensation on the soft palate during the hospital visit were measured, too. Guidance regarding an accepted way of stretch training fitting various disorder parts, such as gum rubbing, the Mendelsohn maneuver, thermal tactile stimulation, the head lift exercise (Shaler exercise), the tongue holding maneuver and Sylvester maneuver. Trainings involved procedures suitable for each patient after obtaining first-person informed consent. For example, patients with the deglutition disorder were performed by thermal tactile stimulation during the hospital visit and Shaker exercise in the residence. Patients performed

Finally, we examined the relationship between the total scores for all items in the oral and pharyngeal phases and the threshold of electrical sensation on the soft palate. We hypothesized that if electrical threshold measurement can be substituted for VF assessments, radiation exposure to patients can be reduced. Furthermore, electrical threshold stimuli on the soft palate and other areas are produced by an appeal of a loose press feeling in the stimulus area of

We aimed to elucidate the relationship between changes in the electrical threshold of sensa‐ tions associated with various orofacial symptoms and the recovery processes of areas affected by orofacial disorders depending on the oral rehabilitation technique. Thus, we investigated the relationship between modified VF assessments and sensory electrical threshold evoked on the soft palate. In particular, electrical stimuli are appealed by a loose press feeling in the

**First experiment:** We examined 11 patients (8 males, 3 females; age range, 30-66 years) with various disorders in orofacial regions (facial muscles, n=4; lingual muscles, n=5; hypoesthesia ofthe soft palate, n=2)(Table 1).All medicaltreatments were performed over a long time period (range, 10-80 months). We compared the relationship between the degree of recovery follow‐ ing oralrehabilitation and the electricalthreshold of sensation in the deficient orofacialregions.

the training method best suited to their disease twice daily.

subjects, not painful sensation.

**2. Material and method**

**2.1. Subjects**

106 Seminars in Dysphagia

stimulus area of subjects, not painful sensation.

**Second experiment:** Eight patients (six males, two females were evaluated; age range, 24-67 years) with glossopharyngeal nerve paralysis of the soft palatal lift. The duration of the patients' dysphasia symptoms ranged from 7 to 39 months before the investigation (Table 2). The eight patients exhibited various symptoms as follows: brain tumor (n=2), cerebral hemorrhage (n=2), cerebral contusion (n=2), amyotrophic lateral sclerosis (n=1), and myasthe‐ nia gravis (n=1) (Table 2). The degree of dysphagia was classified based on many VF assess‐ ments as described by Logemann [8] (underside of Modified VF worksheet, Table 3). The bolus formation, tongue-to-palate contact, premature bolus loss, and oral transit time were measured in the oral phase, while lift in the soft palate, triggering of pharyngeal swallow, epiglottic vallecula residue, pyriform sinus residue, pharyngeal transit time, and aspiration were measured in the pharyngeal phase. Thus, the analysis comprised the total assessments in the oral and pharyngeal phases. These items were divided into three levels: grade 0 as normal, grade 1 as inadequate, and grade 2 as a true abnormality.


**Table 2.** Patient characteristics in the second experiment. NG: no symptoms, PEG: gastrostomy.

We utilized the total scores for all items in the oral and pharyngeal phases separately. In particular, we considered that a higher sum in each phase might indicate a more serious condition. All procedures were approved by the ethic committee at Nihon University School of Dentistry.

#### **2.2. Threshold of electrical sensations**

**2.3. Modified VF worksheet**

**First experiment:** A special electrode was designed for electrical stimulation of the tongue dorsum. This electrode consists of a button-like electrode with a central part (cathode, 0.5 mm across) and round wire (anode, 1.0 mm across) (Figure 1A). This electrode performed the local electrical stimulation. An electrode was applied to the facial skin as a skin patch (one side was the cathode and the other side was the anode) with a 25-mm distance between commercial electrodes (Figure 1B). The electrode applied to the soft palate was a circular disc cathode with a 1.0-cm diameter and circular disc anode with a 1.0-cm diameter made of stainless steel (Figures 1C and 1D). We categorized as "C" or "D" according to various oral conditions. The electrodes were separated according to their use in patients with natural teeth (Figure 1C) versus complete dentures (Figure 1D). The low electrode impedances were approximately 6.5-7.5 kΩ at 20 Hz as measured by Neuropack-μ, Nihon Koden Co.) and the stimulations were 0.2 msec in duration and 5 Hz in frequency.

Figure 1. The electrode was used in various regions: (A) tongue dorsum, (B) facial skin, and (C,D) soft palate. (A) The special button-like electrode consists of a central part (cathode, 0.5 mm across) and round wire (anode, 1.0 mm across); this electrode is suitable for local stimulation. (B) Commercial-release electrodes were used for facial skin stimulation. (C, D) The electrodes were separated according to their use in patients with (C) natural teeth versus (D) complete dentures. Dummy Text **Second experiment:** The electrodes for electrical stimulation were connected to thin acrylic maxillary **Figure 1.** The electrode was used in various regions: (A) tongue dorsum, (B) facial skin, and (C,D) soft palate. (A) The special button-like electrode consists of a central part (cathode, 0.5 mm across) and round wire (anode, 1.0 mm across); this electrode is suitable for local stimulation. (B) Commercial-release electrodes were used for facial skin stimulation. (C, D) The electrodes were separated according to their use in patients with (C) natural teeth versus (D) complete den‐ tures.

splints or full dentures depending on their fit in the paralytic regions of each patient. The stimulus part for electrical stimulation were determined depending on exploration of the hypoesthesia on the soft plate. The electrode comprises a circular disc cathode and anode, both 1.0 cm in diameter, and is made of stainless steel. The electrode mounted at the top of the shielded elastic wire (braided wires) ensured good electrical contact with the moist nature of this part of the mouth. A silicon impression material (GC Co.) was used to cover the surfaces of the stimulating electrodes at low electrode impedances (approximately 6.5-7.5 kΩ at 20 Hz as measured by Neuropack-μ, Nihon Koden Co.). Stöhr and Petruch [12] and Stöhr et al. [13] reported that electrical stimulation in lips or gingival mucosa was appropriate in 0.1-0.2msec (duration) and 1-20 Hz (frequency). On the basis of these data, we decided to stimulate with 0.2msec in duration and 5 Hz in frequency. We employed the first sensation at the increased value with serial access from 0.2 mA. Measurements were performed three times, and the threshold value was averaged. We expected to determine the degree of recovery by the changes in the threshold of electrical sensations. We measured the threshold of electrical sensation in the soft palate with the use of Neuropack-μ every arrival at the hospital. The electrode was jointed with the dental resin on a palatal plate denture or full denture made from each patient, were always stimulated on the same point in every checks.

Although the VF worksheet proposed in Logemann's study [8] is used by many dentists and otolaryngologists, many medical institutions use a more concise version. We divided the dysphasia scale into the oral phase and pharyngeal phase. In the oral phase, we examined the bolus formation, tongue-to palate contact, premature bolus loss, residue in the oral cavity, and oral transit time. In the pharyngeal phase, we examined the lift in the soft palate, triggering of pharyngeal

swallow, epiglottic vallecula, pyriform sinus residue, pharyngeal transit time, and aspiration.

**Second experiment:** The electrodes for electrical stimulation were connected to thin acrylic maxillary splints or full dentures depending on their fit in the paralytic regions of each patient. The stimulus part for electrical stimulation were determined depending on exploration of the hypoesthesia on the soft plate. The electrode comprises a circular disc cathode and anode, both 1.0 cm in diameter, and is made of stainless steel. The electrode mounted at the top of the shielded elastic wire (braided wires) ensured good electrical contact with the moist nature of this part of the mouth. A silicon impression material (GC Co.) was used to cover the surfaces of the stimulating electrodes at low electrode impedances (approximately 6.5-7.5 kΩ at 20 Hz as measured by Neuropack-μ, Nihon Koden Co.). Stöhr and Petruch [12] and Stöhr et al. [13] reported that electrical stimulation in lips or gingival mucosa was appropriate in 0.1-0.2msec (duration) and 1-20 Hz (frequency). On the basis of these data, we decided to stimulate with 0.2msec in duration and 5 Hz in frequency. We employed the first sensation at the increased value with serial access from 0.2 mA. (depending on the mechanical mature).Measurements were performed three times, and the threshold value was averaged. We expected to determine the degree of recovery by the changes in the threshold of electrical sensations. We measured the threshold of electrical sensation in the soft palate with the use of Neuropack-μ every arrival at the hospital. The electrode was jointed with the dental resin on a palatal plate denture or full denture made from each patient, were always stimulated on the same point in every checks. We defined the electrical threshold as the patients were firstly sensed minimum values.

#### **2.3. Modified VF worksheet**

Pa ent No. E ology Loca on Age Sex Time a er onset Status

We utilized the total scores for all items in the oral and pharyngeal phases separately. In particular, we considered that a higher sum in each phase might indicate a more serious condition. All procedures were approved by the ethic committee at Nihon University School

**First experiment:** A special electrode was designed for electrical stimulation of the tongue dorsum. This electrode consists of a button-like electrode with a central part (cathode, 0.5 mm across) and round wire (anode, 1.0 mm across) (Figure 1A). This electrode performed the local electrical stimulation. An electrode was applied to the facial skin as a skin patch (one side was the cathode and the other side was the anode) with a 25-mm distance between commercial electrodes (Figure 1B). The electrode applied to the soft palate was a circular disc cathode with a 1.0-cm diameter and circular disc anode with a 1.0-cm diameter made of stainless steel (Figures 1C and 1D). We categorized as "C" or "D" according to various oral conditions. The electrodes were separated according to their use in patients with natural teeth (Figure 1C) versus complete dentures (Figure 1D). The low electrode impedances were approximately 6.5-7.5 kΩ at 20 Hz as measured by Neuropack-μ, Nihon Koden Co.) and the stimulations were

> Figure 1. The electrode was used in various regions: (A) tongue dorsum, (B) facial skin, and (C,D) soft palate. (A) The special button-like electrode consists of a central part (cathode, 0.5 mm across) and round wire (anode, 1.0 mm across); this electrode is suitable for local stimulation. (B) Commercial-release electrodes were used for facial skin stimulation. (C, D) The electrodes were separated according to

**Figure 1.** The electrode was used in various regions: (A) tongue dorsum, (B) facial skin, and (C,D) soft palate. (A) The special button-like electrode consists of a central part (cathode, 0.5 mm across) and round wire (anode, 1.0 mm across); this electrode is suitable for local stimulation. (B) Commercial-release electrodes were used for facial skin stimulation. (C, D) The electrodes were separated according to their use in patients with (C) natural teeth versus (D) complete den‐

**A B C D**

Dummy Text **Second experiment:** The electrodes for electrical stimulation were connected to thin acrylic maxillary splints or full dentures depending on their fit in the paralytic regions of each patient. The stimulus part for electrical stimulation were determined depending on exploration of the hypoesthesia on the soft plate. The electrode comprises a circular disc cathode and anode, both 1.0 cm in diameter, and is made of stainless steel. The electrode mounted at the top of the shielded elastic wire (braided wires) ensured good electrical contact with the moist nature of this part of the mouth. A silicon impression material (GC Co.) was used to cover the surfaces of the stimulating electrodes at low electrode impedances (approximately 6.5-7.5 kΩ at 20 Hz as measured by Neuropack-μ, Nihon Koden Co.). Stöhr and Petruch [12] and Stöhr et al. [13] reported that electrical stimulation in lips or gingival mucosa was appropriate in 0.1-0.2msec (duration) and 1-20 Hz (frequency). On the basis of these data, we decided to stimulate with 0.2msec in duration and 5 Hz in frequency. We employed the first sensation at the increased value with serial access from 0.2 mA. Measurements were performed three times, and the threshold value was averaged. We expected to determine the degree of recovery by the changes in the threshold of electrical sensations. We measured the threshold of electrical sensation in the soft palate with the use of Neuropack-μ every arrival at the hospital. The electrode was jointed with the dental resin on a palatal plate denture or full denture made from each patient, were always stimulated on the same point in every checks.

Although the VF worksheet proposed in Logemann's study [8] is used by many dentists and otolaryngologists, many medical institutions use a more concise version. We divided the dysphasia scale into the oral phase and pharyngeal phase. In the oral phase, we examined the bolus formation, tongue-to palate contact, premature bolus loss, residue in the oral cavity, and oral transit time. In the pharyngeal phase, we examined the lift in the soft palate, triggering of pharyngeal

swallow, epiglottic vallecula, pyriform sinus residue, pharyngeal transit time, and aspiration.

their use in patients with (C) natural teeth versus (D) complete dentures.

**2.3. Modified VF worksheet**

**Table 2.** Patient characteristics in the second experiment. NG: no symptoms, PEG: gastrostomy.

of Dentistry.

108 Seminars in Dysphagia

tures.

**2.2. Threshold of electrical sensations**

0.2 msec in duration and 5 Hz in frequency.

Although the VF worksheet proposed in Logemann's study [8] is used by many dentists and otolaryngologists, many medical institutions use a more concise version. We divided the dysphasia scale into the oral phase and pharyngeal phase. In the oral phase, we examined the bolus formation, tongue-to palate contact, premature bolus loss, residue in the oral cavity, and oral transit time. In the pharyngeal phase, we examined the lift in the soft palate, triggering of pharyngeal swallow, epiglottic vallecula, pyriform sinus residue, pharyngeal transit time, and aspiration.

For both the oral and pharyngeal phases, the bolus formation, tongue-to-palate contact, premature bolus loss, residue in the oral cavity, oral transit time, lift in the soft palate, and triggering of the pharyngeal swallow were categorized into three levels: grade 0 as normal, grade 1 as inadequate, and grade 2 as a true abnormality (Table 3). In particular, the amount of premature bolus loss, residue in the oral cavity, and epiglottic vallecula and pyriform sinus residue were divided into three levels: grade 0 indicated no residue, grade 1 indicated residue in <50% of the bolus, and grade 2 indicated residue in >50% of the bolus. Furthermore, aspiration was divided into three levels: grade 0 indicated no aspiration, grade 1 indicated supraglottic penetration and grade 2 indicated subglottic aspiration.

All oral and pharyngeal phase criteria were categorized into three levels, 0, 1 and 2. Based on the assortment of the three levels according to the standardized definition, wthat a more serious condition in the oral or pharyngeal phase would be indicated by a higher total for each criterion, as shown in the totals for the oral and pharyngeal phases in Table 3. We considered that assessments performed using these three simplified levels would provide consistency


with respect to the criteria used during long-term evaluation (>2 months after symptom onset). The criterion for each phase (oral and pharyngeal) was employed by the total.

hospital visits. Patients 1-6 had brain disease, and patients 7 and 8 had spinal cord disease and a muscular disorder, respectively.. **Table 3.** Measurements of videofluoroscopic dysphagia scale (VF assessments) and threshold of electrical sensation during the patients' hospital visits. Patients 1-6 had brain disease, and patients 7 and 8 had spinal cord disease and a muscular disorder, respectively..

Table 3. Measurements of videofluoroscopic dysphagia scale (VF assessments) and threshold of electrical sensation during the patients'

For both the oral and pharyngeal phases, the bolus formation, tongue-to-palate contact, premature bolus loss, residue in

#### the oral cavity, oral transit time, lift in the soft palate, and triggering of the pharyngeal swallow were categorized into three levels: grade 0 as normal, grade 1 as inadequate, and grade 2 as a true abnormality (Table 3). In particular, the **3. Results**

**3. Results**

amount of premature bolus loss, residue in the oral cavity, and epiglottic vallecula and pyriform sinus residue were divided into three levels: grade 0 indicated no residue, grade 1 indicated residue in <50% of the bolus, and grade 2 indicated residue in >50% of the bolus. Furthermore, aspiration was divided into three levels: grade 0 indicated no aspiration, grade 1 indicated supraglottic penetration and grade 2 indicated subglottic aspiration. All oral and pharyngeal phase criteria were categorized into three levels, 0, 1 and 2. Based on the assortment of the three levels according to the standardized definition, wthat a more serious condition in the oral or pharyngeal phase would be indicated by a higher total for each criterion, as shown in the totals for the oral and pharyngeal phases in Table 3. We considered that assessments performed using these three simplified levels would provide consistency with respect to the criteria used during long-term evaluation (>2 months after symptom onset). The criterion for each phase (oral and pharyngeal) was employed by the total. **First experiment:** We evaluated the recovery process of each patient as they performed routine exercises and gave an oral assessment of how they felt at each monthly hospital visit. The relationships between the recovery process and the electrical threshold of sensations are indicated by regression curves for each patient (Figures 1A, tongue dorsum; 1B, facial skin; and 1C, soft palate). The regression curves of the electrical threshold and time course showed ① y=-0.0442X+3.1564, r2=0.58404 (acoustic nerve tumor), ② y=-0.0105X+2.425, r2=0.0363 (diabetes), ③ y=-0.0269X+2.4865, r2=0.6517 (cerebral infarct) and ④ y=-0.0177X+4.2982, r2=0.02184 (neck tumor) on tongue dorsum stimulation, ⑤ y=-0.1669X+6.2984, r2=0.22006 (neck tumor), ⑥ y=-0.1605X+3.0904, r2=0.45346 (neck tumor), ⑦ y=-0.0159X+1.3826, r2=0.43144 (cerebral infarct), ⑧ y=-0.0237X+4.4219, r2=0.08506 (neck tumor), ⑨ y=0.0984X+4.1664, r2=0.5461 (facial nerve palsy) on facial skin stimulation, and ⑩ y=-0.0448X+12.404, r2=0.00864 (cerebral infarct), and ⑪ y=-0.0022X+3.5931, r2=3E-05 (cerebral infarct) on soft palate stimula‐ tion. These findings suggested that the electrical threshold stimulation decreased with the time course of the recovery process, because all "gradients" of these data were observed as negative values (Figure 2), although each r2 value (regression estimate of approximate curve) was

Dummy Text **First experiment:** We evaluated the recovery process of each patient as they performed routine exercises and gave an oral assessment of how they felt at each monthly hospital visit. The relationships between the recovery process and the electrical threshold of sensations are indicated by regression curves for each patient (Figures 1A, tongue dorsum; 1B, facial skin; and 1C, soft palate). The regression curves of the electrical threshold and time course showed ① y=-0.0442X+3.1564, r2=0.58404 (acoustic nerve tumor), ② y=-0.0105X+2.425, r2=0.0363 (diabetes), ③ y=-0.0269X+2.4865, r2=0.6517 (cerebral infarct) and ④y=-0.0177X+4.2982, r2=0.02184 (neck tumor) on tongue dorsum stimulation, ⑤ y=-0.1669X+6.2984, r2=0.22006 (neck tumor), ⑥ y=-0.1605X+3.0904, r2=0.45346 (neck tumor), ⑦ y=-0.0159X+1.3826, r2=0.43144 (cerebral infarct), ⑧ y=-0.0237X+4.4219, r2=0.08506 (neck tumor), ⑨ y=0.0984X+4.1664, r2=0.5461 (facial nerve palsy) on facial skin stimulation, and ⑩ y=-0.0448X+12.404, r2=0.00864 (cerebral infarct), and ⑪ y=-0.0022X+3.5931, r2=3E-05 (cerebral infarct) on soft palate stimulation. These findings suggested that the electrical threshold stimulation decreased with the time course of the recovery process, because all "gradients" of these data were observed as negative values (Figure 2), although each r2 value (regression estimate of approximate curve) was indicated by divergence or indicated by divergence or convergence. In particular, although each disease showed a different slope, the recovery trend appeared to be unrelated to the onset of treatment. convergence. In particular, although each disease showed a different slope, the recovery trend appeared to be unrelated

to the onset of treatment.

with respect to the criteria used during long-term evaluation (>2 months after symptom onset).

Table 3. Measurements of videofluoroscopic dysphagia scale (VF assessments) and threshold of electrical sensation during the patients' hospital visits. Patients 1-6 had brain disease, and patients 7 and 8 had spinal cord disease and a muscular disorder, respectively..

**Table 3.** Measurements of videofluoroscopic dysphagia scale (VF assessments) and threshold of electrical sensation during the patients' hospital visits. Patients 1-6 had brain disease, and patients 7 and 8 had spinal cord disease and a

For both the oral and pharyngeal phases, the bolus formation, tongue-to-palate contact, premature bolus loss, residue in the oral cavity, oral transit time, lift in the soft palate, and triggering of the pharyngeal swallow were categorized into three levels: grade 0 as normal, grade 1 as inadequate, and grade 2 as a true abnormality (Table 3). In particular, the amount of premature bolus loss, residue in the oral cavity, and epiglottic vallecula and pyriform sinus residue were divided into three levels: grade 0 indicated no residue, grade 1 indicated residue in <50% of the bolus, and grade 2 indicated residue in >50% of the bolus. Furthermore, aspiration was divided into three levels: grade 0 indicated no

**First experiment:** We evaluated the recovery process of each patient as they performed routine exercises and gave an oral assessment of how they felt at each monthly hospital visit. The relationships between the recovery process and the electrical threshold of sensations are indicated by regression curves for each patient (Figures 1A, tongue dorsum; 1B, facial skin; and 1C, soft palate). The regression curves of the electrical threshold and time course showed ① y=-0.0442X+3.1564, r2=0.58404 (acoustic nerve tumor), ② y=-0.0105X+2.425, r2=0.0363 (diabetes), ③ y=-0.0269X+2.4865, r2=0.6517 (cerebral infarct) and ④ y=-0.0177X+4.2982, r2=0.02184 (neck tumor) on tongue dorsum stimulation, ⑤ y=-0.1669X+6.2984, r2=0.22006 (neck tumor), ⑥ y=-0.1605X+3.0904, r2=0.45346 (neck tumor), ⑦ y=-0.0159X+1.3826, r2=0.43144 (cerebral infarct), ⑧ y=-0.0237X+4.4219, r2=0.08506 (neck tumor), ⑨ y=0.0984X+4.1664, r2=0.5461 (facial nerve palsy) on facial skin stimulation, and ⑩ y=-0.0448X+12.404, r2=0.00864 (cerebral infarct), and ⑪ y=-0.0022X+3.5931, r2=3E-05 (cerebral infarct) on soft palate stimula‐ tion. These findings suggested that the electrical threshold stimulation decreased with the time course of the recovery process, because all "gradients" of these data were observed as negative values (Figure 2), although each r2 value (regression estimate of approximate curve) was

All oral and pharyngeal phase criteria were categorized into three levels, 0, 1 and 2. Based on the assortment of the three levels according to the standardized definition, wthat a more serious condition in the oral or pharyngeal phase would be indicated by a higher total for each criterion, as shown in the totals for the oral and pharyngeal phases in Table 3. We considered that assessments performed using these three simplified levels would provide consistency with respect to the criteria used during long-term evaluation (>2 months after symptom onset). The criterion for each phase (oral and

Dummy Text **First experiment:** We evaluated the recovery process of each patient as they performed routine exercises and gave an oral assessment of how they felt at each monthly hospital visit. The relationships between the recovery process and the electrical threshold of sensations are indicated by regression curves for each patient (Figures 1A, tongue dorsum; 1B, facial skin; and 1C, soft palate). The regression curves of the electrical threshold and time course showed ① y=-0.0442X+3.1564, r2=0.58404 (acoustic nerve tumor), ② y=-0.0105X+2.425, r2=0.0363 (diabetes), ③ y=-0.0269X+2.4865, r2=0.6517 (cerebral infarct) and ④y=-0.0177X+4.2982, r2=0.02184 (neck tumor) on tongue dorsum stimulation, ⑤ y=-0.1669X+6.2984, r2=0.22006 (neck tumor), ⑥ y=-0.1605X+3.0904, r2=0.45346 (neck tumor), ⑦ y=-0.0159X+1.3826, r2=0.43144 (cerebral infarct), ⑧ y=-0.0237X+4.4219, r2=0.08506 (neck tumor), ⑨ y=0.0984X+4.1664, r2=0.5461 (facial nerve palsy) on facial skin stimulation, and ⑩ y=-0.0448X+12.404, r2=0.00864 (cerebral infarct), and ⑪ y=-0.0022X+3.5931, r2=3E-05 (cerebral infarct) on soft palate stimulation. These findings suggested that the electrical threshold stimulation decreased with the time course of the recovery process, because all "gradients" of these data were observed as negative values (Figure 2), although each r2 value (regression estimate of approximate curve) was indicated by divergence or

aspiration, grade 1 indicated supraglottic penetration and grade 2 indicated subglottic aspiration.

pharyngeal) was employed by the total.

muscular disorder, respectively..

**3. Results**

110 Seminars in Dysphagia

**3. Results**

The criterion for each phase (oral and pharyngeal) was employed by the total.

Figure 2. Changes in the recovery process of threshold electrical stimulation. Each start point indicated the start of treatment after onset of the disease. Dummy Text **Second experiment:** All patients in the present study have severe swallowing dysfunctions as shown in **Figure 2.** Changes in the recovery process of threshold electrical stimulation. Each start point indicated the start of treatment after onset of the disease.

Table 1. These disorders, which share characteristics in common with dysphasia, were divided into two types depending on the underlying cause. The location of the underlying cause was classified as either the brain (patients 1-6) or spinal cord (patients 7 and 8) (Table 2). All criteria of the oral and pharyngeal phases were categorized into three levels: 0, 1 and 2 (Table 2). We investigated the relationship between the threshold of sensation and each value in the VF assessment of each patient's disease condition. **Second experiment:** All patients in the present study had severe swallowing dysfunctions as shown in Table 1. These disorders, which share characteristics in common with dysphasia, were divided into two types depending on the underlying cause. The location of the under‐ lying cause was classified as either the brain (patients 1-6) or spinal cord (patients 7 and 8) (Table 2).

These values were grouped into classes according to the data shown in Table 3. The total score of the VF assessments in the oral and pharyngeal phases and the threshold of sensation in each patient are shown in Figures 3Aa, 3Ab and 3B; the averages and standard deviations in each figure are indicated in Figures 3C and 3D. Almost all data showed a recovery trend (negative "gradient" of the regression curve) with the exception of amyotrophia lateral sclerosis and myasthenia gravis. Thus, amyotrophic lateral sclerosis and myasthenia gravis did not exert a curative influence on our treatment, because suggestion based on the VF assessments and electrical threshold of sensation did not recover (Figure 3: black lines, amyotrophic lateral sclerosis; red lines, myasthenia gravis). These results may suggest that such muscle diseases do not recovery with the type of training performed in this study. Although the average scores of the VF assessments and threshold electrical stimulation indicated a recovery trend for all diseases, separate analysis for each disease was importance to obtain more detailed data. Based on these findings, we believe that the electrical threshold of sensation on the soft plate may reflect the recovery process of swallowing reflex disorders. All criteria of the oral and pharyngeal phases were categorized into three levels: 0, 1 and 2 (Table 2). We investigated the relationship between the threshold of sensation and each value in the VF assessment of each patient's disease condition. These values were grouped into classes according to the data shown in Table 3. The total score of the VF assessments in the oral and pharyngeal phases and the threshold of sensation in each patient are shown in Figures 3Aa, 3Ab and 3B; the averages and standard deviations in each figure are indicated in Figures 3C and 3D. Almost all data showed a recovery trend (negative "gradient" of the regression curve) with the exception of amyotrophia lateral sclerosis and myasthenia gravis. Thus, amyotrophic lateral sclerosis and myasthenia gravis did not exert a curative influence on our treatment, because suggesting based on the VF and threshold that sensation did not recover (Figure 3: black lines, amyotrophic lateral sclerosis; red lines, myasthenia gravis). These results may suggest that such muscle diseases do not recovery with the type of training performed in this study. Although the average scores of the VF assessments and threshold electrical stimulation indicated a recovery trend for all diseases, separate analysis for each disease was importance to obtain more detailed data. Based on these findings, we believe that the electrical threshold of sensation on the soft plate may reflect the recovery process of swallowing reflex disorders.

**Figure 3.** Relationship between total scores of VF assessments in (A) the oral phase and (D) the pharyngeal phase and the threshold of electrical sensation. (Aa) VF assessments in the oral phase. (Ab) VF assessments in the pharyngeal phase. (B) Electrical threshold of sensation. Black lines: amyotrophic lateral sclerosis. Red lines: myasthenia gravis. (C, D) Averages and standard deviations of VF assessments and electrical threshold of sensation.
