unc speech lab

Pressure-Flow Laboratory

UNC Craniofacial Center

David J. Zajac, Ph.D.

Associate Professor/Director

The Pressure-Flow Laboratory provides aerodynamic assessment of patients with craniofacial anomalies, mostly repaired cleft lip and/or palate. The lab originated as a result of the pioneering work of Don Warren, DDS, PhD, in the 1960s. Today, the lab continues to serve as a clinical and research arm of the UNC Craniofacial Center. The primary function of the lab is the assessment of the nasal airway and velopharyngeal structures during respiration and speech in individuals with various craniofacial disorders. In addition, the lab has the capacity to provide Nasometric, electroglottographic (EGG), and acoustic evaluations of patients.

The UNC Pressure-Flow Laboratory

Lab Personnel

Ø David J. Zajac, PhD, CCC-SLP Director

Ø Lynn Fox, MEd, CCC-SLP Craniofacial Center SLP

Ø Fernando Querios Doctoral Student

Clinical Function of the Lab

As a clinical arm of the Craniofacial Center, lab personnel assess patients who are seen for new and follow-up diagnostic team evaluations. Two primary assessments are performed: determination of the patency of the nasal airway during breathing and adequacy of velopharyngeal structures during speech production. Children as young as 3 years of age often can be assessed (see modifications to pressure-flow technique below). Typically, children who are 4 years of age and older are evaluated on a routine basis. During 2006, over 130 pressure-flow evaluations were performed.

Why Use Aerodynamic Assessment?

Individuals with velopharyngeal dysfunction typically exhibit speech characteristics that include hypernasality, weak pressure consonants, audible nasal air emission, and compensatory articulations. Characteristics such as hypernasality, weak pressure consonants, and nasal air emission are considered obligatory because these will occur to various degrees whenever there is incomplete separation of the oral and nasal cavities by the velopharyngeal structures. While hypernasality is primarily an acoustic-perceptual phenomenon, weak pressure consonants and nasal air emission are primarily aerodynamic in nature. The perceptual descriptions of these symptoms vary widely among clinicians, especially descriptions of nasal air emission. Aerodynamic assessment of the magnitude of oral and nasal air pressures and of nasal air loss, therefore, is essential if one wishes to objectively evaluate the extent of velopharyngeal dysfunction during speech production. Some compensatory articulations such as a co-produced mid-dorsum palatal stop/posterior nasal fricative can also be readily assessed and differentiated from obligatory symptoms by an experienced clinician who uses pressure-flow techniques.

Finally, given that the goal of cleft palate surgery is to repair the velopharyngeal structures in order for the child to achieve normal speech production, surgeons should be especially interested in pressure-flow techniques. As described below, pressure-flow techniques can objectively document if the primary goal of surgery (i.e., achieving normal velopharyngeal closure) has been achieved.

Aerodynamic Principles and Techniques

The following is a brief tutorial that explains basic principles and techniques of aerodynamic assessment. A list of selected references follows the tutorial.

Pressure-Flow Technique

The pressure-flow technique was first applied to assess velopharyngeal (VP) function during speech by Warren and colleagues during the 1960s. Based upon principles of fluid mechanics, the technique enables an estimate of the size of any existing VP gap during speech production. As illustrated below, the size of an orifice can be calculated if the differential pressure and volume rate of flow can be determined.

In the above illustration, water flows through a pipe and encounters a constriction. In order to flow through the constriction, there is an increase in volume velocity with a subsequent loss of driving pressure (illustrated by the change in height of the water column downstream from the constriction). The area of the constriction (or orifice) can be calculated by using a derivative of Bernoulli’s equation:

Area = (V/√ΔP) (.11)

where V is the volume rate of flow and ΔP is the differential pressure. The constant (.11) includes corrections for density of the fluid and a discharge coefficient that is determined by the shape of the orifice. When applied to speech, the “orifice equation” will estimate VP area in square millimeters (mm²) if nasal airflow (V) is expressed in ml/s and differential oral-nasal air pressure (ΔP) is expressed in cm H₂O.

Application of the above principles to the human vocal tract is illustrated below.

In this photograph, oral air pressure is tapped by a small plastic tube placed in the mouth behind the lips. The oral air pressure represents the upstream pressure before the constriction in the pipe illustration. A second plastic tube is embedded in a foam plug and inserted in one of the nostrils. This seals the nostril and taps nasal air pressure. The nasal air pressure represents the downstream pressure after the constriction in the pipe illustration. Finally, a flow tube is fitted into the remaining nostril to channel nasal airflow. The oral and nasal pressure tubes are attached to separate differential pressure transducers referenced to atmosphere. The nasal flow tube is connected in series to a heated pneumotachograph and differential pressure transducer (see equipment table in lab photograph above). We use commercially-available hardware and software (PERCI-SARS, Microtronics, Chapel Hill, NC ) to acquire and display the aerodynamic speech data. In addition, a microphone records the acoustic speech signal.

Because VP closure is complete during production of the voiceless stop-plosive /p/ by speakers without cleft palate, this speech segment is often used to assess VP function. In our laboratory, we obtain pressure-flow data using the following speech samples. Click on each sample below to see a pressure-flow output from a non-cleft speaker (nasal air pressure is omitted from the output in the examples).

· The sentence “Peep into the hamper”

Each of the syllables and the word “hamper” are produced approximately 5 times as a continuous utterance (i.e., a single breath group). The sentence is produced 3 times also on a single breath group. Because some children with VP dysfunction may reduce loudness as a compensatory strategy to mask obligatory symptoms (i.e., “soft voice syndrome”), we instruct all children to use their habitual loudness levels. If a child uses substantially reduced loudness, then we will ask the child to repeat the sample using a strong voice.

Zajac (2000), referenced below, provides normative pressure-flow data for children and adults without cleft palate producing the above speech samples.

Determination of Simultaneous Oral and Nasal Airflow

A limitation of the pressure-flow approach as presented above is that velopharyngeal (VP) orifice size is most reliably estimated during production of consonants which have relatively high differential oral-nasal pressure (i.e., oral pressure consonants). During production of vowels, the oral cavity is open and differential pressure is negligible. If an assessment of VP function is desired during production of vowels and low pressure consonants, then the clinician may want to determine simultaneous oral and nasal airflow.

The photograph above illustrates the use of an oral-nasal circumferentially-vented pneumotachograph mask (Glottal Enterprises, Syracuse, NY ). The mask is partitioned into nasal and oral chambers. Each chamber has fine mesh, wire screens that serve as flow-resistive pneumotachographs. Catheters are inserted into each chamber to detect pressure variations associated with airflow. A microphone (not visible in the photograph) is positioned outside of the mask to record the audio signal.

While estimation of VP orifice size using the pressure-flow approach is independent of respiratory effort and/or oral cavity configuration (i.e., tongue placement of vowels), the magnitude of nasal airflow is influenced by both of these factors. Accordingly, a ratio of nasal to oral-plus-nasal airflow is determined to index VP function. [This is similar to the calculation of acoustic nasalance.] In addition, integration of the airflow signals is often done to determine lung volumes associated with speech segments and/or utterances (see graphic #1 below). In Graphic #1, the speaker exhibits severe VP dysfunction and expends more nasal air volume than oral air volume during the utterance.

Click here to view Oral-Nasal Mask graphic #1

As a comparison to the pressure-flow approach, the speech samples illustrated below were obtained using an oral-nasal mask. The same non-cleft speaker produced both sets of samples.

· The sentence “Peep into the hamper”

Assessment of Velopharyngeal Timing

Aerodynamic techniques are ideally suited to assess timing aspects of velopharyngeal (VP) function. As suggested by Warren and colleagues, temporal aspects of opening and closing gestures may be as important (or more so in some cases) than the actual size of the VP gap. Duration of the VP closing gesture can be determined from nasal airflow data obtained during production of speech samples containing nasal-plosive sequences as in the word “hamper” illustrated below.

The above output is from a non-cleft speaker who produced the word “hamper.” Nasal airflow is segmented as follows: A to B reflects anticipatory nasal airflow prior to the /m/ segment, B to C reflects the beginning to peak nasal airflow associated with the /mp/ sequence, and C to E reflects peak to the end of nasal airflow associated with the /mp/ sequence. Given the nasal-plosive context, the VP closing gesture can be inferred as the time from C to E (approximately 75 ms in this example). As noted by Dotevall et al. (2002), nasal airflow may not decline to zero if a speaker has VP dysfunction. Because of this, Dotevall et al. (2002) suggest that the end of nasal airflow be considered at 5% to baseline. In the above example, the horizontal dotted line indicates the 5% to baseline airflow criterion. Finally, it may be necessary to normalize the VP closing duration if comparisons are to be made across speakers who differ in speaking rate (e.g., adults and children). This can be done by using the duration of the nasal airflow segment from A to E. This segment reflects approximately the duration of the first syllable of the word “hamper.”

Dotevell et al. (2002) also suggested that the negative peak of the derivative of nasal airflow (Point D above) may be a useful index of the VP closing gesture. The negative peak reflects the fastest rate of nasal airflow declination. Dotevall et al. (2002) normalized this value by reference to peak nasal airflow (Point C above).

The diagnostic importance of VP timing measures was highlighted by Warren et al. (1993). They reported that speakers with repaired cleft palate and adequate VP closure who were judged to be hypernasal demonstrated nasal airflow durations (segment B to E above) that were approximately 50 ms longer than speakers with repaired cleft palate and adequate VP closure who were judged to be hypernasal and controls.

Determination of Laryngeal Airway Resistance

Because many speakers with VP dysfunction also present with perceptual symptoms of vocal dysfunction and/or laryngeal anomalies, assessment of laryngeal airway resistance (LAR) is a valuable clinical tool. LAR is defined as the quotient of the transglottal pressure difference to oral airflow during vowel production. To determine LAR, a plastic catheter is inserted through the oral chamber of the oral-nasal mask and positioned in the mouth to detect oral air pressure (see graphic #2). The speaker is instructed to repeat the syllable /pi/ seven times.

Click here to view Oral-Nasal Mask graphic #2

Myth Busting

As with any instrumental assessment technique, limitations exist. I have described some of these in the above tutorial. In this section, I will present some common myths and discuss modifications to procedures that further extend the application of aerodynamic techniques.

Myth: The pressure-flow technique is limited to assessment of the voiceless bilabial stop /p/.

If one wishes to estimate VP orifice size, then this is most easily done by targeting the /p/ segment. This phoneme is ideal because a) obstruent consonants in general are produced with high differential oral-nasal air pressure, and b) oral air pressure is most easily tapped during production of bilabial stops. To detect oral air pressure during a stop consonant, the open end of the catheter must be positioned behind the constriction. This is easily achieved for a bilabial stop, somewhat easy to achieve for an alveolar stop, but rather difficult to achieve for a velar stop. To tap pressure during a velar stop, either a buccal-sulcus or transnasal approach to catheter placement must be used. Obviously, for clinical purposes, this is not often done. Because airflow ceases during stop consonants, orientation of the open end of the catheter is unimportant.

As illustrated above for the speech sample /si/, the pressure-flow approach can also be used to assess fricatives. When detecting oral air pressure for a fricative, however, orientation of the catheter is important. Because airflow continues, the end of the catheter must be positioned perpendicular to the flow, or, the end of the catheter must be occluded and side holes provided in the wall of the catheter. We use the former approach and position the catheter behind the tongue from the corner of the lips, thus achieving a perpendicular orientation to the flow.

Aerodynamic procedures provide much more than just estimates of VP orifice size. Indeed, as illustrated above, important information regarding respiratory breath groups and temporal aspects of VP function are provided. Furthermore, this information can be obtained from inspection of nasal airflow data without the need to obtain differential pressure.

Myth: The pressure-flow technique cannot be used to assess continuous speech.

This myth is similar to the above criticism. Obviously, by constructing the appropriate speech samples, continuous speech can be assessed. In our protocol, only continuous speech is assessed because we instruct the speaker to produce syllables and words within a single breath group. Because of this, we can also assess declination (or fatigue) effects across the utterance if these occur. We also instruct speakers to produce the sentence stimulus (“peep into the hamper”) by taking a relatively deep breath and saying all three repetitions on a single breath group. This instruction provides information on the effects of respiratory effort in addition to possible fatigue. We often observe that VP orifice size during production of /p/ in “peep into the hamper” will be reduced compared to “hamper” at word level because of increased respiratory effort (i.e., the speaker initiates the sentence at a higher lung volume level). Our point is that careful construction of speech samples may provide important diagnostic information.

Myth: The pressure-flow technique cannot be used with very young children.

Although cooperation is an important issue, especially if one wants to determine VP orifice size, modifications to the pressure-flow technique can be done to further extend its application to children as young as 3 years of age. In our lab, we use two modifications.

1. Use of a nasal mask instead of flow tube. This modification reduces the level of cooperation required of the child by eliminating multiple items (flow tubes and pressure taps) inserted into the nose. Instead, nasal mask pressure is tapped as nasal airflow is obtained. This modification, however, requires a correction of the differential pressure. To do this, nasal resistance during breathing must also be determined. This procedure and correction is described in Zajac (2005). The use of only a nasal flow mask and no oral catheter may also be used. Obviously, this modification will only provide nasal airflow information.

Illustration of nasal mask modification

2. Use of differential oral-nasal pressure. This modification eliminates the need to obtain nasal airflow. Use of oral and nasal pressure taps are still required and the speech sample should include “hamper.” Evaluation of VP function is based upon data suggested by Warren (1997): differential pressure greater than 3.0 cm H₂O is considered adequate while differential pressure less than 1.0 cm H₂O is considered inadequate.

Illustration of differential pressure modification

Myth: Because VP orifice estimates are two-dimensional, the information provided is not important.

Obviously, no single instrumental technique provides all the necessary information for diagnosis and/or treatment. Proponents of this myth further argue that a single number cannot be used to represent a complex phenomenon such as speech. Indeed, a single number cannot. If the goal of cleft palate surgery, however, is to establish normal VP closure during speech, then pressure-flow estimates of orifice size are extremely valuable. Furthermore, as I have attempted to illustrate above, aerodynamic techniques provide much more than just VP orifice estimates and do so in non-invasive and minimally intrusive ways.

Selected References

Basic Principles and Early Studies

Lubker, J.F., & Moll, K.L. (1965). Simultaneous oral-nasal airflow measurements and cineflurorgraphic observations during speech production. Cleft Palate J. 2: 257-272.

Smitheran, R., & Hixon, T.J. (1981). A clinical method for estimating laryngeal airway resistance during vowel production. J Speech Hear. Disord. 46:138-146.

Warren, D.W., & DuBois, A. (1964). A pressure-flow technique for measuring velopharyngeal orifice area during continuous speech. Cleft Palate Journal, 1, 52-71.

Warren, D.W. (1984). A quantitative technique for assessing nasal airway impairment. American Journal of Orthodontics, 86, 306-314.

Yates, C.C., McWilliams, B.J., & Vallino, L.D. (1990). The pressure-flow method: Some fundamental concepts. Cleft Palate Journal, 27, 193-198.

Zajac, D.J., & Yates, C.C. (1991). Accuracy of the pressure-flow method in estimating induced velopharyngeal orifice area: Effects of the flow coefficient. Journal of Speech and Hearing Research, 34, 1073-1078.

Timing Studies of VP Function

Dotevall, D.R., Lohmander-Agerskov, A., Ejnell, H., & Bake, B. (2002). Perceptual evaluation of speech and velopharyngeal function in children with and without cleft palate and the relationship to nasal airflow patterns. Cleft Palate-Craniofacial Journal, 39: 409-424.

Warren, D.W., Dalston, R.M., & Mayo, R. (1993). Hypernasality in the presence of "adequate" velopharyngeal closure. Cleft Palate J. 30:150-154.

Warren D.W., Dalston, R.M. Trier, W.C., & Holder, M.B. (1985). A Pressure-Flow technique for quantifying temporal patterns of palatopharyngeal closure. Cleft Palate Journal, 22, 11-19.

Zajac, D., & Hackett, A. (2002). Temporal characteristics of aerodynamic segments in the speech of children and adults. Cleft Palate-Craniofacial Journal, 39:432-438. 3

Zajac, D.J. & Mayo R. (1996). Aerodynamic and temporal aspects of velopharyngeal function in normal speakers. Journal of Speech and Hearing Research, 39, 1199-1207.

Clinical Applications and Normative Data

Thompson, A.E., & Hixon, T.J. (1979). Nasal airflow during normal speech production. Cleft Palate Journal, 16:412-420.

Warren, D.W. (1997). Aerodynamic assessments and procedures to determine extent of velopharyngeal inadequacy. In: K.R. Bzoch (ed.), Communicative Disorders Related to Cleft Lip and Palate, Fourth Edition. Austin Texas: PRO-ED, Inc.

Zajac, D. (2000). Pressure-flow characteristics of /m/ and /p/ production in speakers without cleft palate: Developmental findings. Cleft Palate-Craniofacial Journal, 37:468-477. 3

Zajac, D. (2005). Maximizing Clinical Acquisition and Interpretation of Aerodynamic and Acoustic Speech Data in Children with Velopharyngeal Dysfunction. In “Perspectives,” ASHA Special Interest Division 5 Newsletter, 15(2): 11-14.

to David Zajac's homepage Last Updated: 10/19/10