The influence of acoustic and perceptual factors on perceived hypernasality in the vowel

The purpose of this study was to examine the influence of acoustic and perceptual factors of speech on listeners' perceived hypernasality in the vowel [i]. The isolated syllable [pi] produced by 22 children with hypernasal speech and 6 noncleft children was rated by 10 listeners. These speech samples were then divided into two groups: (1) the samples (n = 14) that received inconsistent ratings from each listener or variable ratings among listeners (i.e., unreliable ratings) and (2) the samples (n = 14) that received consistent ratings from each listener and similar ratings among listeners (i.e., reliable ratings). These results suggest that the severity of hypernasality was easy to rate in some speech samples and not in others. Voice quality deviation and a particular type of spectral change that related to the severity of hypernasality could be factors that influence perceived hypernasality.

The relationship between spectral characteristics and perceived hypernasality in children

The purpose of this study was to quantify perceived hypernasality in children. One-third octave spectra of the isolated vowel [i] were obtained from 32 children with cleft palate and 5 children without cleft palate. Four experienced listeners rated the severity of hypernasality of the 37 speech samples using a 6-point equal-appearing interval scale. When the average 1/3-octave spectra from the hypernasal group and the normal resonance group were compared, spectral characteristics of hypernasality were identified as increased amplitudes between F1 and F2 and decreased amplitudes in the region of F2. Based on the findings of the children's speech, 36 speech samples with manipulated spectral characteristics were used to minimize the influences of voice source characteristics on perceived hypernasality. Multiple regression analysis revealed a high correlation (R = 0.84) between the amplitudes of 1/3-octave bands (1 k, 1.6 k, and 2.5 kHz) and the perceptual ratings. Increased amplitudes of bands between F1 and F2 (1 k, 1.6 kHz) and decreased amplitude of the band of F2 (2.5 kHz) was associated with an increasing perceived hypernasality. These results suggest that the amplitudes of the three 1/3-octave bands are appropriate acoustic parameters to quantify hypernasality in the isolated vowel [i].

Assessment of aspiration risk in stroke patients with quantification of voluntary cough

BACKGROUND

Dysphagia and subsequent aspiration are serious complications of acute stroke that may be related to an impaired cough reflex. It was hypothesized that aspirating stroke patients would have impaired objective measures of voluntary cough as compared with both nonstroke control subjects and nonaspirating stroke patients.

METHODS

Swallowing was evaluated by standard radiologic or endoscopic methods, and stroke patients were grouped by aspiration severity (severe, n = 11; mild, n = 17; no aspiration, n = 15). Airflow patterns and sound pressure level (SPL) of voluntary cough were measured in stroke patients and in a group of normal control subjects (n = 18). Initial stroke severity was determined retrospectively with the Canadian Neurological Scale.

RESULTS

All cough measures were altered in stroke patients as a group relative to nonstroke control subjects. Univariate analysis showed that peak flow of the inspiration phase (770.6 +/- 80.6 versus 1,120.1 +/- 148.4 mL/s), SPL (90.0 +/- 3.1 versus 100.2 +/- 1.6 dB), peak flow of the expulsive phase (875.1 +/- 122.7 versus 1,884.1 +/- 221.6 mL/s), expulsive phase rise time (0.34 +/- 0.1 versus 0.09 +/- 0.01 s), and cough volume acceleration (5.5 +/- 1.3 versus 27.8 +/- 3.9 mL/s/s) were significantly impaired in severe aspirators as compared with nonaspirators. Aspirating patients had more severe strokes than nonaspirators (mean Canadian Neurological Scale score 7.7 +/- 0.7 versus 9.8 +/- 0.3). Multivariate logistic regression found only expulsive phase rise time values during cough correlated with aspiration status.

CONCLUSION

Objective analysis of cough may provide a noninvasive way to identify the aspiration risk of stroke patients.

Effects of vowel height and vocal intensity on anticipatory nasal airflow in individuals with normal speech

The purpose of this study was to determine the effects of vowel height and vocal intensity on the magnitude of anticipatory nasal airflow in normal speakers when producing vowel-nasal-vowel (VNV) sequences. Measurements of nasal and oral airflow were obtained from 15 men and 12 women with normal speech during production of the VNV sequences /ini/ and /ana/ at low, medium, and high intensity levels. Ratios of nasal to oral-plus-nasal airflow were calculated for the initial vowel of both utterances at each of the intensity levels. Analysis of variance (ANOVA) procedures indicated a significant main effect of intensity level and a significant vowel-by-sex interaction effect (p < .05) on the airflow ratios. Overall, the airflow ratio was reduced at high as compared to low intensity levels, regardless of sex of the speaker or vowel type. Female speakers exhibited greater airflow ratios during production of /ini/ than during productions of /ana/. Their airflow ratios were also greater during production of /ini/ than were those of male speakers. The results suggest that vocal intensity may affect velopharyngeal (VP) function in an assimilative nasal phonetic context. The results further suggest that anticipatory nasal airflow may be determined by the configuration of the oral cavity to a greater extent in women than in men. Theoretical and clinical implications are discussed.

Pressure-flow characteristics of /m/ and /p/ production in speakers without cleft palate: developmental findings

OBJECTIVE

The purpose of this study was to describe the pressure-flow characteristics of a large sample of speakers without cleft palate ranging in age from early childhood to young adulthood.

METHOD

Speakers consisted of 223 children, teens, and adults without cleft palate categorized into five age groups: 6 through 8 years, 9 through 10, 11 through 12, 13 through 16, and 18 through 37 years. Speakers produced the syllables /mi/, /pi/, and /p/¿/, the word "hamper," and the sentence "Peep into the hamper." The pressure-flow method was used to determine oral air pressure, nasal airflow, and estimates of velopharyngeal (VP) orifice size associated with /ml and /p/ production. Descriptive statistics were computed for each age group and speech sample. Analysis of variance (ANOVA) procedures were used to determine the effects of age, sex, and production level (word versus sentence) on the aerodynamic variables.

RESULTS

ANOVA procedures indicated significant main effects (p < .01) of age on most of the aerodynamic variables during production of /m/ and /p/. No significant main effects or interactions involving sex were found for any variable. Regardless of age, approximately 95% to 99% of the speakers exhibited airtight VP closure during /p/ at syllable level, depending upon the selected nasal airflow criterion. ANOVA procedures also indicated significant main effects of production level (word versus sentence) on each of the aerodynamic variables during the /mp/ sequence. These effects appeared to be related to speaking rate.

CONCLUSIONS

The study suggests that speakers without cleft palate exhibit essentially complete VP closure during production of oral pressure consonants in isolated syllables, and developmental aspects of speech aerodynamics be considered during pressure-flow testing.

Aerodynamic characteristics of tracheostomy speaking valves: an updated report

Two analyses were conducted to define some aerodynamic properties of one-way speaking valves designed for use with the tracheotomized patient. In the first analysis, the resistance to airflow of six different valves was determined during steady-state flow testing at rates of .450, .500, and .550 I/s. Significant differences among the valves were established only at the lowest flow rate. All valves exhibited relatively low resistance in the range of nasal resistance reported for normal adults. In the second analysis, the aerodynamic integrity of the valves was assessed during repetition of the syllable /pa/ under a condition used to simulate tracheostomy speech production. Significant differences were found among the valves in terms of air loss occurring during the rise in pressure associated with the production of the consonant /p/. Valves with diaphragms open at atmospheric pressure consistently exhibited air loss. Average slope of the rise in pressure for one of the valves tested was significantly greater, suggesting increased work during speech production. The results of these analyses suggest that although the inspiratory resistance to airflow was similar among various one-way speaking valves, some valves exhibit air loss during speech production.

Intraoral pressure and velopharyngeal function

OBJECTIVE

The objective of this study was to determine the influence of velopharyngeal (VP) inadequacy on respiratory speech compensations.

DESIGN

The pressure-flow technique was used to measure pressure, airflow, and timing variables associated with VP closure during the production of the initial plosive consonant /p/ in a series of the utterance "papa."

SETTING

The study was conducted in the speech and breathing laboratory of the UNC Craniofacial Center.

PARTICIPANTS

Eighty-two subjects with cleft lip and/or palate were assessed. The subjects were divided into two groups, those with adequate VP closure (VP size <.010 cm2) and those with inadequate VP closure (VP size >0.10 cm2). The adequate group was comprised of 62 subjects, and 20 subjects were categorized as inadequate.

RESULTS

Peak intraoral pressure decreased in the inadequate group, but the difference was not significant. Nasal airflow increased (p < .01), but duration of the pressure pulse was the same for both groups. The area under the pressure curve decreased for the inadequate group (p = .04).

CONCLUSION

These data contrast with previously reported published data using /p/ in the utterance "hamper." This suggests that phonetic context influences the compensatory response to velopharyngeal inadequacy. Additionally, while the findings are somewhat similar to studies that involved noncleft subjects whose oral airway was suddenly vented during the production of /p/, there is enough difference to suggest that learning also affects the compensatory outcome.

Nasal coarticulation in normal speakers: a re-examination of the effects of gender

The purpose of this study was to re-examine the influence of gender on nasal coarticulation in normal speakers. Twenty adult speakers (10 men, 10 women) produced the vowel-nasal-vowel (VNV) sequence /ini/ within a carrier phrase using two stress patterns: (1) with equal stress placed on both syllables, and (2) with contrastive stress placed on the second syllable. A partitioned, circumferentially vented pneumotachograph mask and microphone were used to determine nasal airflow, ratios of nasal to oral-plus-nasal airflow, and sound pressure levels (SPLs) at the midpoints of the first and second vowels of the syllables. Volume ratios of nasal to oral-plus-nasal airflow that reflected the entire duration of the vowels were also obtained. Results indicated that all speakers except 1 woman exhibited nasal airflow at the midpoint of the first vowel when /ini/ was produced with equal stress; all speakers exhibited carryover nasal airflow during the second vowel. During contrastive stress, all speakers except 1 woman and 1 additional man exhibited anticipatory nasal airflow; all speakers exhibited carryover nasal airflow. Analysis of variance (ANOVA) procedures with repeated measures indicated significant main effects of syllable stress (p < 0.001) relative to nasal airflow and ratios of nasal to oral-plus-nasal airflow. Regardless of the gender of the speakers, syllable stress resulted in reduced anticipatory and carryover nasal airflow during the first and second vowels of /ini/, respectively. The results suggest that (a) both men and women adhere to similar patterns of velar articulation, and (b) velopharyngeal closure during the vowel /i/ may be enhanced during stressed syllables. Implications relative to aspects of speech production and clinical practice are discussed.

Effects of a pressure target on laryngeal airway resistance in children

The purpose of this study was to determine the effects of an intraoral air pressure target on estimation of laryngeal airway resistance (LAR) in normal children. Ten children produced the syllable /pi/ (a) with self-determined normal loudness, (b) with increased loudness, and (c) at a predetermined intraoral air pressure level of 6.5 to 7.5 cm of water. The target pressure level was selected because it was expected to result in estimated subglottal pressures that were lower than those associated with self-determined loudness levels. Results indicated significant differences in estimated subglottal pressure among the three conditions. As expected, estimated subglottal pressures were highest during loud speech and lowest during the pressure target task. LAR values associated with normal loudness were similar to values previously reported for children. The use of the pressure target resulted in LAR values that were reduced by 31% from normal loudness. These resistance values, however, were still greater than those reported for adult speakers at similar subglottal pressure levels. The results are explained relative to preferred loudness levels and vocal tract size differences in children and adults. It is suggested that use of a pressure target during estimation of LAR in children may provide additional data that more accurately reflect the aerodynamic integrity of the larynx. Implications for clinical assessment are discussed.

Velopharyngeal function in young and older adult speakers: evidence from aerodynamic studies

Intraoral air pressure (P0), nasal airflow, and nasal air volume characteristics were determined from 63 adults speakers during production of oral and nasal consonants. The speakers were categorized into two age groups. Young speakers (21 males, 21 females) ranged in age from 18 to 37 years (mean = 24); older speakers (10 males, 11 females) ranged in age from 68 to 83 years (mean = 75). Speakers produced the syllables /pi/ and /si/ and the word "hamper" at self-determined loudness levels during reiterant speech tasks. Results indicated similar velopharyngeal function between young and older adults as reflected by nasal airflow and air volume measures. Older adults, however, exhibited significantly higher levels of P0 than younger adults during production of /p/ in "hamper." In addition (a) male speakers exhibited higher P0 than females during production of /p/ in "hamper," (b) all speakers demonstrated declination of P0 during production of /p/ in "hamper" as a function of word position in the utterance, and (c) male and female speakers exhibited different patterns of nasal air volume during production of "hamper" as a function of word position in the utterance. The intraoral air pressure results suggest differences in respiratory function between young and older speakers. The nasal air volume results suggest a gender specific velopharyngeal declination effect. Implications for velar motor control strategies are discussed.

The response to sudden change in vocal tract resistance during stop consonant production

The speech respiratory system is configured in ways that tend to maximize its ability to respond to changes in the airway environment. Intraoral pressures remain at levels sufficient to generate reliably recognized consonant sounds even in the presence of structural deficits such as velopharyngeal inadequacy. Similar respiratory compensations occur when bite blocks and bleed valves are used to vent airway pressures. The purpose of the present study was to determine the sensitivity of the monitoring system psychophysically and to assess its physiological response to sudden, unanticipated perturbations. Twenty adults were asked to produce the utterance/pa/, and a calibrated perturbator valve permitted air to escape from the oral cavity on randomly selected productions. Respiratory responses were recorded using PERCI-SARS instrumentation. The results indicated that sudden openings of 0.14 cm2 (SD = 0.04) were detected by speakers. Compensatory respiratory responses to suprathreshold pressure-venting occurred rapidly (i.e., 27 ms [SD = 8]) after valve opening. Although peak pressure and area under the pressure pulse fell with valve opening, the magnitude of pressure was nevertheless sufficient for sound generation. Measurements of the slope of the rise in intraoral pressure after subthreshold pressure-venting in 10 participants were compared to measurements obtained from an-elastic model of the upper airway. The data demonstrated a significant difference between vented and unvented conditions for the model, but not the participants. This suggests that elastic recoil is actively and unconsciously controlled in humans to compensate for losses in airway pressure during speech.

Microphone sensitivity as a source of variation in nasalance scores

A two-part study was conducted to determine the sources of variation in nasalance scores derived from the Nasometer. In Study #1, a function generator was used as a signal source to calibrate and input sine and square waves directly into the Nasometer. Ten stimuli ranging from 105 to 330 Hz in 25 Hz increments were evaluated. In Study #2, the same signal source and an amplified loudspeaker were used to calibrate and present square waves to the nasometer via five different sets of microphones. The sound pressure level of all stimuli was maintained at 88 dB. Each microphone set was calibrated using the 105 Hz signals. Results from Study #1 indicated consistent nasalance scores across all frequencies (i.e., all scores were within 2% of calibration). Results from Study #2 demonstrated deviations greater than 2% from calibration as a function of frequency for all five sets of microphones. The smallest deviation was 5%, whereas the largest deviation was 14%. We suggest that the variation in nasalance as a function of stimulus frequency may be due to a mismatch in the sensitivity of microphones (i.e., different frequency response characteristics). It is further suggested (a) that individual investigators determine the response characteristics of their microphones and (b) that relatively small variations in nasalance scores (i.e., 5-14%) either within or across speakers be interpreted with caution.

Aerodynamic and temporal aspects of velopharyngeal function in normal speakers

Aerodynamic and temporal characteristics of velopharyngeal function were determined for 42 adult male and female speakers. All subjects produced the word "hamper" at self-determined loudness levels and rates of speaking. Measurements of intraoral air pressure, nasal airflow, and estimates of velopharyngeal orifice size were obtained during production of the /m/ and /p/ segments. Volume measurements of nasal airflow were determined for the entire word, the /m/ segment, and the segments preceding /m/. Fifteen timing measures associated with the pressure-flow events of the nasal-plosive sequence were also determined. Results indicated that males generated significantly higher levels of peak intraoral air pressure than females during /p/. Male speakers also exhibited a significantly shorter interval in the rise of oral pressure associated with the /p/ segment. Male and female speakers, however, exhibited similar levels of anticipatory coarticulation as reflected by nasal air volume measurements. Finally, variability of selected measurements within speakers suggested that temporal aspects of velopharyngeal function were more constrained than aerodynamic aspects. The results are discussed relative to (a) sex differences in respiratory and velar function and (b) normative data for adult speakers.

Aerodynamic and acoustic characteristics of a speaker with turbulent nasal emission: a case report

Aerodynamic and acoustic characteristics were determined from the speech of an adult female with mild mental retardation and severe velopharyngeal inadequacy. The speaker's productions of /s/ were characterized by consistent nasal grimacing and turbulent air emission. Aerodynamic assessment estimated the size of the velopharyngeal orifice to exceed 200 mm2 during plosive production. Nasal cross-sectional area was estimated to be 35 mm2 during quiet breathing. Nasometric evaluation indicated nasalance of 63% associated with the "Zoo" passage. Acoustic analysis of the separately recorded oral and nasal speech signals indicated spectral energies in the region of approximately 2.5 to 7.0 kHz associated with nasal emission during /s/ production. The occurrence of these frequencies suggested an acoustic/perceptual function of the nasal grimace. Pressure-flow evidence also suggested that the nasal grimace, perhaps with lingual assistance, functioned to enhance speech aerodynamics.

Dyspnea following experimentally induced increased nasal airway resistance

Nasal resistance (NRZ) values for healthy adults range from 1.0 to 3.5 cm H2O/L/sec. Some oral breathing tends to occur at values above 3.5. The purpose of the present study was to determine at what level of NRZ individuals sense that nasal breathing is difficult. A diaphragm was used to add four different resistance loads in random to 15 adult subjects. These loads were 5,8, and 15 cm H2O/L/sec and a value 40% above the individual's normal NRZ. Loads were added under four conditions: normal breathing, fixed flow rate, fixed breathing rate, and fixed flow and breathing rate. The pressure-flow technique was used to measure NRZ under all conditions. The study revealed that the sensation of breathing difficulty occurred at a median resistance of 5 cm H2O/L/sec and, as subjects were constrained to maintain fixed flow and breathing rates, the magnitude of RZ, at which the sensation of dyspnea was noted, decreased. The values observed in this study support previous findings suggesting that individuals switch to some oral breathing to maintain an adequate level of upper airway resistance at values between 3.5 and 4.5 cm H2O/L/sec. The findings also show that individuals attempt to minimize increases in airway resistance by modifying breathing behaviors.

Laryngeal airway resistance in children with cleft palate and adequate velopharyngeal function

Laryngeal airway resistance during vowel production was determined for 10 children without cleft palate and 14 children with cleft palate and adequate velopharyngeal function. The children with cleft palate were further grouped according to either complete or incomplete velopharyngeal closure. All children performed a syllable repetition task with nostrils occluded at self-determined effort and a targeted effort typical of adult speech. Results indicated that all children exhibited significantly greater laryngeal resistance at self-determined effort. No significant differences occurred among the groups of children at either effort level. Some children with incomplete velopharyngeal closure, however, tended to exhibit relatively increased laryngeal resistance at the targeted effort level when their nostrils were unoccluded. Implications for a regulation-control model of speech production and clinical assessment are discussed.

Aerodynamic characteristics of tracheostomy speaking valves

Pressure-flow characteristics were determined for four different one-way valves used for speech production in the patient who has had a tracheotomy. Each valve was tested at steady-state flow rates of 150, 250, 350, and 450 ml/sec in isolation and attached to a tracheostomy tube. Results indicated significant differences in resistance among the valves. The resistance of one valve was substantially greater than that of the normal upper airways. It is suggested that future research determine the clinical significance, if any, of these differences relative to speech and respiratory behaviors in tracheostomized patients.

Age and gender effects on nasal respiratory function in normal subjects

One hundred and ninety-seven normal individuals between the ages of 5 and 73 years were evaluated to determine nasal resistance, nasal cross-sectional area, and respiratory mode during quiet breathing. Subjects were categorized into three age groups. Nasal resistance and respiratory mode were directly determined using posterior rhinomanometry and the SNORT technique, respectively. Nasal cross-sectional area was estimated using the hydrokinetic equation. Results indicated significant effects of age on all variables; significant gender differences were found for respiratory mode. Weak correlations were found between respiratory mode and nasal resistance. The results are presented as normative data on nasorespiratory characteristics to facilitate diagnostic and treatment decisions relative to individuals with normal morphology as well as to patients with craniofacial anomalies. A fundamental issue of both clinical and theoretical importance arising from the study pertains to the definitions of normality and impairment.

Voice perturbations of children with perceived nasality and hoarseness

Cycle-to-cycle variations in voice fundamental frequency (jitter) and amplitude (shimmer) were derived by electroglottography for 10 children with velopharyngeal insufficiency (VPI). Jitter was found to be positively correlated with ratings of perceived nasality, whereas shimmer was found to be positively correlated with ratings of perceived hoarseness. Theoretic implications for a regulatory model of speech aerodynamics are discussed. Additionally, clinical applications, in terms of using electroglottography as a supplemental assessment procedure, are suggested.