Swallow Sounds Research Articles

Characterization of Masticatory and Swallowing Sounds Using Wavelet Analysis

Characterization of Masticatory and Swallowing Sounds Using Wavelet Analysis

Acoustic characteristics associated with liquid swallowing sounds of different bolus consistencies in young healthy adults.

Cervical auscultation has been used clinically as an augmentative procedure for swallow examination. Despite its frequent use for screening and preliminary assessment of swallowing, the usefulness of cervical auscultation is controversial due to a lack of sufficient research, particularly in quantifying swallowing sounds. To document the acoustic characteristics of normal swallowing sound associated with swallowing bolus of different consistencies among younger healthy adults. A total of 30 healthy young adult participants swallowed 5 ml thickened liquids of four different consistencies (slightly thick, mildly thick, moderately thick and extremely thick) prepared using a starch-based commercial thickener, and the corresponding swallowing sounds were recorded using a wireless accelerometric stethoscope. An array of acoustic measures including duration of swallowing sound (DSS), duration to peak intensity (DPI), frequency at peak intensity (FPI), peak intensity (PI), average intensity (AI), and difference between peak and average intensity (DPAI) values associated with the swallowing signals were obtained. In general, increased durational measures of the swallowing sounds were associated with an increase in bolus consistency. Intensity measures including PI, AI and DPAI were found to be stable across different consistencies. The change in FPI did not appear to be particularly meaningful due to its high variability. In addition, no significant differences were observed between men and women. Swallow sounds associated with different bolus consistencies could be quantified and used to differentiate consistencies. The present findings could serve as a reference for future swallowing research of normal and dysphagic population. What is already known on the subject Cervical auscultation using traditional stethoscope has been used as part of an informal clinical swallow examination by practitioners. Validity of cervical auscultation is controversial, possibly due to the lack of normative data on swallow sounds. What this paper adds to existing knowledge The present study explored the possibility of using wireless accelerometric stethoscopy for cervical auscultation for dysphagia screening. Acoustic profiles of swallow sounds associated with boluses of different consistencies in healthy individuals were examined. What are the potential or actual clinical implications of this work? Findings contribute to our knowledge about the acoustic characteristics of swallow sounds of boluses of different consistencies in healthy young individuals. The study provides normative clinical data on cervical auscultation using wireless accelerometric stethoscope for normal swallow.

Dysphagia management: Does structured training improve the validity and reliability of cervical auscultation?

Purpose: Cervical auscultation (CA) uses a stethoscope or microphone to complement the clinical swallow examination by interpreting swallowing sounds and swallow-respiratory coordination. This study investigated the effects of structured CA training on CA-rating agreement with Flexible Endoscopic Evaluation of Swallowing (FEES) and CA rater reliability. Method: Thirty-nine speech-language pathologists participated in a structured CA training course at Gothenburg University. They rated nine swallow-respiratory sound recordings which were simultaneously recorded during FEES. Swallow sounds were rated six weeks prior to the CA-workshop using two binary yes/no questions, (1) Safe, (2) Dysphagia, and a third Dysphagia Severity rating. Swallow sounds were rated again (re-randomised) one month post CA-workshop. Result: Agreement with FEES (validity) improved significantly (p < 0.05) pre-post training for the Safe and Dysphagia questions, with post training sensitivities >90% and specificities at 76% and 85% respectively. Dysphagia severity rating improved non-significantly. Intra-rater reliability improved significantly with kappa statistics >0.90 post training. Improvements for inter-rater reliability were noted, though non-significant. Conclusion: Results demonstrate that with structured training, the validity of CA (to detect a Safe/Dysphagic swallow) significantly improves, as does intra-rater reliability. This is congruent with literature identifying the positive effects of structured training improving instrumental dysphagia assessment.

The effects of neuromuscular joint facilitation treatment using neck patterns on the acoustic analysis of swallowing sounds in healthy adults.

[Purpose] We aimed to investigate the changes in the swallowing sounds of healthy adults during neuromuscular joint facilitation treatment using neck patterns. [Participants and Methods] A total of 20 healthy adults (10 males and 10 females; mean age, 29.2 ± 6.3 years) swallowed 10 mL of water four times under three conditions (after the neuromuscular joint facilitation neck-flexion resistance pattern, after the Shaker-type exercise, and during relaxed sitting without prior exercise [control]), randomly ordered with an interval greater than 3 days. Swallowing sounds for each water swallow were recorded using cervical auscultation. [Results] The mean amplitude of swallowing sound intensity and the mean spectral frequency were significantly higher after the neuromuscular joint facilitation neck-flexion resistance pattern and the Shaker-type exercise, in comparison with those in the control group. [Conclusion] Neuromuscular joint facilitation training with the neck-flexion resistance pattern influenced swallowing sounds to the same degree as the Shaker-type exercise, implying that this resistance pattern may enhance suprahyoid muscle contraction.

Deep Learning-Based Swallowing Monitor for Realtime Detection of Swallow Duration.

Aspiration pneumonia is a life-threatening disease for the elderly. To prevent its risk, regular swallowing assessment is necessary; however, current screening tools for swallow assessment are not widely available and medical experts are insufficient. As a portable assessment tool, we have been developing a smartphone-based realtime monitoring device (GOKURI) which can evaluate swallowing ability based on swallow sounds. For better detection accuracy of the system, we integrated a deep learning model which was developed based on the swallowing anatomy. In this paper, we provide a detailed analysis to see how the swallow sounds detected by the deep learning-based monitor correspond to the actual swallow activities. Also, as an example of practical application of the system, we analyzed the changes of the swallow abilities over time by recording swallow sounds twice for the same participants at a nursing home. To minimize the risk of aspiration pneumonia, caregivers need to understand the disability levels of the patient's swallows so that safe feeding assistance can be provided. The result of this paper implies the possibility of using GOKURI as a daily swallowing monitor with minimum interventions.

Classifying Dysphagic Swallowing Sounds with Support Vector Machines.

Swallowing sounds from cervical auscultation include information related to the swallowing function. Several studies have been conducted on the screening tests of dysphagia. The literature shows a significant difference between the characteristics of swallowing sounds obtained from different subjects (e.g., healthy and dysphagic subjects; young and old adults). These studies demonstrate the usefulness of swallowing sounds during dysphagic screening. However, the degree of classification for dysphagia based on swallowing sounds has not been thoroughly studied. In this study, we investigate the use of machine learning for classifying swallowing sounds into various types, such as normal swallowing or mild, moderate, and severe dysphagia. In particular, swallowing sounds were recorded from patients with dysphagia. Support vector machines (SVMs) were trained using some features extracted from the obtained swallowing sounds. Moreover, the accuracy of the classification of swallowing sounds using the trained SVMs was evaluated via cross-validation techniques. In the two-class scenario, wherein the swallowing sounds were divided into two categories (viz. normal and dysphagic subjects), the maximum F-measure was 78.9%. In the four-class scenario, where the swallowing sounds were divided into four categories (viz. normal subject, and mild, moderate, and severe dysphagic subjects), the F-measure values for the classes were 65.6%, 53.1%, 51.1%, and 37.1%, respectively.

Swallowing Sound Recognition at Home Using GMM

BackgroundAiming for autonomous living for the people after a stroke is the challenge these days especially for swallowing disorders or dysphagia. The most common cause of dysphagia is stroke. In France, stroke occurs every 4 minutes, which implies 13000 hospitalizations per year. Currently, continuous medical home monitoring of patients is not available. The patient must be hospitalized or visit the medical community for possible follow-up. It is in this context that E-SwallHome (Swallowing & Breathing: Modelling and e-Health at Home) project proposes to develop tools, from hospital care until the patient returns home, which are able to monitor in real time the process of swallowing. MethodThis paper presents a relevant health problem affecting patient recovering from stroke. We propose a frequency acoustical analysis for automatic detection of swallowing process and a non-invasive acoustic based method to differentiate between swallowing sounds and other sounds in normal ambient environment during food intake. ResultThe proposal algorithm for events detection gives a global rate of good detection of 87.31%. Classification of sounds of swallowing and other sounds based on Gaussian Mixture Models (GMM), using the leave-one-out approach according to the small amount of data in our database, gives a good recognition rate of swallowing sounds of 84.57%. ConclusionThe proposal method has great potential to assist in the clinical evaluation using only swallowing sounds, which is a non-invasive technic for swallowing studies.

Using Machine Learning and a Combination of Respiratory Flow, Laryngeal Motion, and Swallowing Sounds to Classify Safe and Unsafe Swallowing.

The aim of this research was to develop a swallowing assessment method to help prevent aspiration pneumonia. The method uses simple sensors to monitor swallowing function during an individual's daily life. The key characteristics of our proposed method are as follows. First, we assess swallowing function by using respiratory flow, laryngeal motion, and swallowing sound signals recorded by simple sensors. Second, we classify whether the recorded signals correspond to healthy subjects or patients with dysphagia. Finally, we analyze the recorded signals using both a feature extraction method (linear predictive coding) and a machine learning method (support vector machine). Based on our experimental results for 140 healthy subjects (54.5 32.5 years old) and 52 patients with dysphagia (75.5 20.5 years old), our proposed method could achieve 82.4% sensitivity and 86.0% specificity. Although 20% of testing sample sets were erroneously classified, we conclude that our proposed method may facilitate screening examinations of swallowing function. In combination with the portable sensors, our proposed method is worth utilizing for noninvasive swallowing assessment.

Comparative analysis of acoustic features of spontaneous swallowing sounds between the young and the elderly

Objective To observe the features of spontaneous swallowing sounds of both the young and the elderly. Methods One hundred healthy individuals were randomly divided into an elderly group (60-85 years old) and a youth group (19-30 years old), each of 50 (25 male and female individuals). Both groups were measured with regard to their neck circumference (neck circumference around the thyroid cartilage level) and body mass index (BMI). Spontaneous swallowing sounds were detected for 20 minutes on the thyroid cartilage below the left side of trachea. By using RavenPro1.5.0, the swallowing sounds were segmented, tagged and calculated in terms of the spontaneous swallowing frequency per minute (SPM), swallowing duration, energy consumption, duration of half energy consumption and proportion of half energy consumption duration. All data were finally analyzed using R3.4.0. Results There was no significant difference in the average neck circumference between the youth group and the elderly group [(34.31±3.45) cm versus (33.95±3.00)cm] (P=0.5789). The BMI of the youth group was significantly lower than that of the elderly group [(21.97±2.65)kg/m2 vs (23.43±2.89)kg/m2]. The SPM of the elderly group was significantly lower than that of the youth group [(0.39±0.34) /min vs (0.91±0.51)/min, P<0.0001]. The length of swallowing, the energy consumption per swallowing and the duration of half energy consumption of the elderly group (0.92±0.29 seconds, 76.88±9.82 dB and 0.43±0.24 seconds) were significantly higher than those of the youth group (0.84±0.26 seconds, 75.43±8.93 dB and 0.38±0.19 seconds). However, no significant differences were found in the proportion of half energy consumption duration in the whole swallowing sound duration (P=0.6859). Conclusions Compared with the young, the frequency of spontaneous swallowing decreased and the swallowing duration prolonged significantly in the elderly. Key words: Swallowing; Aging; Acoustic analysis; Spontaneous swallowing frequency; Presbyphagia

Accuracy of cervical auscultation in detecting the presence of material in the airway.

Several studies have investigated the accuracy of cervical auscultation (CA). However, both the sensitivities and the specificities of CA in detecting dysphagic conditions varied widely among these studies. These wide variations of the accuracy of CA might be caused by differences of the targeted sounds, such as the expiratory sound (ES) and/or swallowing sound (SS). Forty‐six dysphagic patients were served as subjects. Patients who had unoccluded tracheostoma and patients who could not follow the instructions were excluded. During the videofluorographic swallowing study (VFSS), the subjects swallowed 3 ml of yogurt containing barium sulfate. The VFSS images were recorded with acoustic signals including both the swallowing and respiratory sounds detected by our method. Classification of the VFSS images was decided by consensus of the three dentists using a penetration‐aspiration scale (PAS). Recorded VFSS images were classified into the following 2 groups based on PAS: “no or minimum risk group”: PAS1–2; and “possible risk group”: PAS3–8. As a result of the classification of VFSS findings, 30 samples were evaluated as no or minimum risk group and 16 as possible risk group. Twelve observers including 10 dentists other than 3 dentists who evaluated VFSS images and 2 speech pathologists auditorily diagnosed “negative” and “positive.” Sensitivity, specificity, and intra‐rater reliability was calculated for the 3 types of acoustic samples. The sensitivity of the intra‐rater reliability was 60.3% for ES, 76.6% for SS, and 89.9% for ES + SS. The sensitivity of intra‐rater reliability of ES + SS samples was significantly higher than that of ES (p < .01) and SS (p < .05). The sensitivity of intra‐rater reliability of SS was significantly higher than that of ES (p < .01). The specificity of the intra‐rater reliability was 53.7% for ES, 50.3% for SS, and 44.5% for ES + SS. ES + SS might be most useful for detecting the presence of material in the airway.

Characteristics of saliva swallowing in the reading of prepared texts

This study examines the nature of saliva swallowing during speech production. Speakers read two contrasting types of texts, a “paragraph” and a “list.” The “paragraph” was a Japanese version of “The North Wind and the Sun,” consisting of four sub-paragraphs. The “list” was a simple repetition of quasi-identical short sentences, without any paragraph construction. Simultaneous recordings of speech sounds (via microphone), swallow sounds (via laryngeal contact microphone), and respiratory movements of the thorax and abdomen (via respiratory belt transducers) were obtained and analyzed, especially in relation to sentence-final pauses. Preliminary results indicate that (1) swallowing tends to occur roughly in the middle of a pause and is followed by inspiration for the upcoming utterance, (2) inspiration is generally deeper when preceded by swallowing, (3) swallowing adds approximately one second to a pause, and (4) in paragraph reading, swallowing is more likely to occur during a pause between two sub-paragraphs, i.e., at a topic transition, whereas in list reading, swallowing reflects speakers' physiological needs more directly.

Aspirating and Nonaspirating Swallow Sounds in Children

Cervical auscultation (CA) may be used to complement feeding/swallowing evaluations when assessing for aspiration. There are no published pediatric studies that compare the properties of sounds between aspirating and nonaspirating swallows. To establish acoustic and perceptual profiles of aspirating and nonaspirating swallow sounds and determine if a difference exists between these 2 swallowing types. Aspiration sound clips were obtained from recordings using CA simultaneously undertaken with videofluoroscopic swallow study. Aspiration was determined using the Penetration-Aspiration Scale. The presence of perceptual swallow/breath parameters was rated by 2 speech pathologists who were blinded to the type of swallow. Acoustic data between groups were compared using Mann Whitney U-tests, while perceptual differences were determined by a test of 2 proportions. Combinations of perceptual parameters of 50 swallows (27 aspiration, 23 no aspiration) from 47 children (57% male) were statistically analyzed using area under a receiver operating characteristic (aROC), sensitivity, specificity, and positive and negative predictive values to determine predictors of aspirating swallows. The combination of post-swallow presence of wet breathing and wheeze and absence of GRS and normal breathing was the best predictor of aspiration (aROC = 0.82, 95% CI, 0.70-0.94). There were no significant differences between these 2 swallow types for peak frequency, duration, and peak amplitude. Our pilot study has shown that certain characteristics of swallow obtained using CA may be useful in the prediction of aspiration. However, further research comparing the acoustic swallowing sound profiles of normal children to children with dysphagia (who are aspirating) on a larger scale is required.

Design of a Signal Conditioning Device for Remote Breath and Swallowing Sounds Recording

Design of a Signal Conditioning Device for Remote Breath and Swallowing Sounds Recording

Characteristics of the swallowing sounds recorded in the ear, nose and on trachea

The various malfunctions and difficulties of the swallowing mechanism necessitate various diagnostic techniques to address those problems. Swallowing sounds recorded from the trachea have been suggested as a noninvasive method of swallowing assessment. However, acquiring signals from the trachea can be difficult for those with loose skin. The objective of this pilot study was to explore the viability of using the ear and nose as alternative recording locations for recording swallowing sounds. We recorded the swallowing and breathing sounds of five healthy young individuals from the ear, nose and trachea, simultaneously. We computed time-frequency features and compared them for the different locations of recording. The features included the peak and the maximum frequencies of the power spectrum density, average power at different frequency bands and the wavelet coefficients. The average power calculated over the 4 octave bands between 150 and 2,400 Hz showed a consistent trend with less than 20 dB difference for the breath sounds of all the recording locations. Thus, analyzing breath sounds recorded from the ear and nose for the purpose of aspiration detection would give similar results to those from tracheal recordings; thus, ear and nose recording may be a viable alternative when tracheal recording is not possible.

Normal Swallowing Acoustics Across Age, Gender, Bolus Viscosity, and Bolus Volume

Cervical auscultation has been proposed as an augmentative procedure for the subjective clinical swallowing examination due to the tangible differences between normal and dysphagic swallowing sounds. However, the research is incomplete regarding cervical auscultation and swallowing acoustics in that the differences between the sounds of normal versus dysphagic swallowing have yet to be fully understood or quantified. The swallows of 96 reportedly healthy adults, balanced for gender and divided into younger, middle, and older age groups, were audio-recorded while ingesting several boluses of varying viscosity and volume. The audio signals were then analyzed to determine their temporal and acoustic characteristics. Results indicated increasing pharyngeal swallowing duration with increasing age, bolus viscosity, and bolus volume. In addition, an increased duration to peak intensity with increasing age was found in one of our two analyses, as well as with some of the more viscous versus less viscous boluses. Men and older persons produced higher peak intensities and peak frequencies than women and younger persons. Thin liquids were produced with more intensity than honey or more viscous boluses, and with greater frequency than mechanical soft solids. Larger volumes resulted in greater peak frequency values. Some of the acoustic measurements appear to be more useful than others, including the duration of the acoustic swallowing signal and the within-subjects peak intensity variable. We noted that differences in swallowing acoustics were more related to changes in viscosity rather than volume. Finally, within-participant observations were more useful than between-participant observations.

Denoising swallowing sound to improve the evaluator’s qualitative analysis

Swallowing dynamics involves the coordination and interaction of several muscles and nerves which allow correct food transport from mouth to stomach without laryngotracheal penetration or aspiration. Clinical swallowing assessment depends on the evaluator’s knowledge of anatomic structures and of neurophysiological processes involved in swallowing. Any alteration in those steps is denominated oropharyngeal dysphagia, which may have many causes, such as neurological or mechanical disorders. Videofluoroscopy of swallowing is presently considered to be the best exam to objectively assess the dynamics of swallowing, but the exam needs to be conducted under certain restrictions, due to patient’s exposure to radiation, which limits periodical repetition for monitoring swallowing therapy. Another method, called cervical auscultation, is a promising new diagnostic tool for the assessment of swallowing disorders. The potential to diagnose dysphagia in a noninvasive manner by assessing the sounds of swallowing is a highly attractive option for the dysphagia clinician. Even so, the captured sound has an amount of noise, which can hamper the evaluator’s decision. In that way, the present paper proposes the use of a filter to improve the quality of audible sound and facilitate the perception of examination. The wavelet denoising approach is used to decompose the noisy signal. The signal to noise ratio was evaluated to demonstrate the quantitative results of the proposed methodology.

Cervical Auscultation Synchronized with Images from Endoscopy Swallow Evaluations

Cervical auscultation is the use of a listening device, typically a stethoscope in clinical practice, to assess swallow sounds and by some definitions airway sounds. Judgments are then made on the normality or degree of impairment of the sounds. Listeners interpret the sounds and suggest what might be happening with the swallow or causing impairment. A major criticism of cervical auscultation is that there is no evidence on what causes the sounds or whether the sounds correspond to physiologically important, health-threatening events. We sought to determine in healthy volunteers (1) if a definitive set of swallow sounds could be identified, (2) the order in which swallow sounds and physiologic events occur, and (3) if swallow sounds could be matched to the observed physiologic events. Swallow sounds were computer recorded via a Littmann stethoscope from 19 healthy volunteers (8 males, 11 females, age range = 18-73 years) during simultaneous fiberoptic laryngoscopy and respiration monitoring. Six sound components could be distinguished but none of these occurred in all swallows. There was a wide spread and a large degree of overlap of the timings of swallow sounds and physiologic events. No individual sound component was consistently associated with a physiologic event, which is a clinically significant finding. Comparisons of groups of sounds and events suggest associations between the preclick and the onset of apnea; the preclick and the start of epiglottic excursion; the click and the epiglottis returning to rest; the click and the end of the swallow apnea. There is no evidence of a causal link. The absence of a swallow sound in itself is not a definite sign of pathologic swallowing, but a repeated abnormal pattern may indicate impairment. At present there is no robust evidence that cervical auscultation of swallow sounds should be adopted in routine clinical practice. There are no data to support the inclusion of the technique into clinical guidelines or management protocols. More evaluation using imaging methods such as videofluoroscopy is required before this subjective technique is validated for clinical use by those assessing swallowing outside of a research context.

False alarms and begging in nestling birds

In theory, for any communication system, receivers cannot increase their responsiveness to signals without simultaneously increasing their risk of responding to inappropriate stimuli. This inevitable trade-off may be an important selective pressure on signalling systems, but has rarely been studied. Nestling tree swallows, Tachycineta bicolor , produce a begging display when their parents arrive at the nest with food, but they also beg to apparently inappropriate stimuli in the absence of parents, such as movements of the nest or broodmates. To explore the implications of these errors for begging, we experimentally determined how begging errors vary with nestling hunger and age, two factors that affect begging rate and intensity in response to parents. Specifically, we examined how begging rate and intensity varied in response to recordings of (1) a tree swallow adult landing on a nestbox and calling and (2) a common grackle, Quiscalis quiscala , a nest predator, landing on a nestbox, during a 1-h period of food deprivation and also across two ages (5–6 and 7–9 days posthatch). Nestlings increased the rate and intensity of their begging responses to both swallow and grackle stimulus sounds as time without food increased, although responses to the grackle sounds were always less than to the swallow sounds. Begging rate, but not intensity, increased with age for both swallow and grackle sounds; both measures of response were lower for the grackle sound at both ages. Thus, as nestlings become more responsive to parents, they also risk committing more errors. This trade-off supports theoretical predictions and may be an important selective force shaping parent–offspring signalling.

Reliability and Validity of Cervical Auscultation: A Controlled Comparison Using Video fluoroscopy

Cervical auscultation is experiencing a renaissance as an adjunct to the clinical swallowing assessment. It is a controversial technique with a small evidence base. We have aimed to establish whether cervical auscultation interpretation is based on the actual sounds heard or, in practice, influenced by information gleaned from other aspects of the clinical assessment, medical notes, or previous knowledge. We sought to determine (a) rater reliability and its impact on the clinical value of cervical auscultation and (b) how judgments compare with the "gold standard": videofluoroscopy. Swallow sounds were computer recorded via a Littmann stethoscope. Sounds were sampled from 10 healthy control swallows with no aspiration/penetration and 10 patient swallows with aspiration/penetration, all recorded during simultaneous videofluoroscopy. The system generated sound quality similar to "live" bedside listening, a feature rarely seen in cervical auscultation studies. The 20 sound clips were classified as "normal" or "abnormal" by 19 volunteer speech-language pathologists with experience in cervical auscultation. After at least four weeks, 11 of these judges rated the sounds rerandomized on a new CD. Intrarater reliability kappa ranged from -0.12 to 0.71. Individual reliability did not correlate with years of experience, practice pattern, or frequency of use. Interrater reliability kappa = 0.17. Comparison with radiologically defined aspiration/penetration yielded 66% specificity, 62% sensitivity, and majority consensus gave 90% specificity, 80% sensitivity. There was a significant relationship between individual reliability and true positive rate (r(s) = 0.623, p = 0.040). The reliability of individual judges varied widely and thus, inevitably, agreement between judges was poor. Validity is dependent upon reliability: Improving the poor raters would improve the overall accuracy of this technique in predicting abnormality in swallowing. The group consensus correctly identified 17 of the 20 clips so we may speculate that the swallow sound contains audible cues that should in principle permit reliable classification.

Inter- and intra-rater reliability of cervical auscultation to detect aspiration in patients with dysphagia.

To measure the inter- and intra-rater reliability of cervical auscultation used alone to detect aspiration in dysphagic patients. A university teaching hospital. Comparison of the detection of aspiration in 16 recorded swallow sounds by five speech and language therapists on two occasions. Swallow sounds were recorded simultaneously with videofluoroscopy. Sixteen patients referred for assessment of dysphagia with videofluoroscopy. The kappa statistic for multiple raters showed fair agreement between raters (kappa = 0.28). There was high agreement when aspiration occurred but in non-aspirating swallows there was significant overdetection of aspiration (p < 0.001 McNemar's test). The intra-rater reliability within different individuals was widely variable (kappa = 0.55 (range 0.31-0.85)). Presented with the swallowing sounds in isolation speech and language therapists cannot reliably classify swallows into those with accompanying aspiration and those without. There appears to be a problem of over detection of aspiration. Even in this small study, however, some individual therapists achieve such high reliability (kappa = 0.85) that they must be using successful internal criteria to interpret the swallow sounds correctly and further qualitative research may identify these.