Академический Документы
Профессиональный Документы
Культура Документы
Summary: Objective. This study compares the flow resistance of Resonance tube (RT) and Lax Vox tube (LVT)
when submerged 2 cm and 10 cm in water, as well as phonation into the tubes in these conditions.
Methods. In the in vitro experiment, the air pressure for flow rates of 60–600 mL/s was measured at the tube inlet,
when the outer end of the tube was submerged 2 cm and 10 cm below water surface in 30°, 45°, and 90° angle. In the
in vivo experiment, 14 subjects phonated in habitual loudness and loudly into both tubes, with the outer end 2 cm and
10 cm in water. RT was immersed in a 45° angle and LVT in a 90° angle in water. Oral pressure, contact quotient from
electroglottographic signal, and sound pressure level were studied. Sensations during phonation were reported in an
interview.
Results. Flow resistance was slightly lower with LVT than with RT, and slightly lower for smaller immersion angles.
In habitual loudness, transglottic pressure and frequency of oral pressure variation were lower for LVT phonation and
amplitude of oral pressure variation was higher for LVT 2 cm in water. Some subjects preferred RT, whereas others
preferred LVT or reported no differences between them.
Conclusions. The tubes differed slightly in flow resistance. Higher oral pressure oscillation with LVT 2 cm in water
may offer stronger massage effect on vocal folds.
Key Words: Flow resistance–EGG–Voice therapy–Water resistance therapy–Air pressure in phonation.
FIGURE 3. Example of oral pressure measurement during tube phonation and random shuttering of the distal end of the tube.
P(trans) was measured as it is the driving force of vocal fold Population variance was investigated by Kolmogorov-Smirnov
vibration. Analyses were made with Soundswell Signal Work- one-sample t test. The sample types (Resonance tube vs Lax Vox
station (Nyvalla DSP, Stockholm, Sweden). Additionally, the tube) were compared using Student’s t test or Wilcoxon signed
frequency and amplitude of peak-to-peak oscillation in P(oral) rank paired t test, depending on the population variance. A non-
during water bubbling were measured (Figure 4). For that purpose, parametric test (Wilcoxon) was used when the values did not
the P(oral) signal was filtered by fourth-order Butterworth digital show normal distribution. The level of statistical significance was
filter in order to exclude acoustic pressure variations from the set to P < 0.05. The SPSS-21 software (International Business
signal. Limits in filtering (cutoff frequencies) were set so that Machines (IBM) Corporation, Armonk, North Castle, New York,
they could reliably exclude both static part (high-pass filter) and the USA) was used in the analyses.
fundamental frequency of vocal fold vibration (low-pass filter)
in all signals. The low limit in filtering was set to 4 Hz and the RESULTS
high limit was set to 30 Hz. Filtering and measurements were Differences between the tubes for the flow
made with Matlab (MathWorks, Natick, MA). The SPL was mea- resistance measured in vitro
sured from the acoustic signal using the Praat software (Paul Figure 6A shows the measured backpressure as a function of
Boersma and David Weenink, version 5.3.55, Amsterdam, The flow rate for both tubes. The differences between the tubes were
Netherlands). Contact quotient (CQ = contact time/period time), small. In the range of 0–0.3 L/s flow rate, which corresponds
derived from EGG, was calculated as it reflects phonation type.15,16 to ordinary phonation,14 the maximum difference considering all
VoceVista (Roden, The Netherlands) was used for automatic cal- immersion angles was only 0.4 cm H2O, which is within the
culation of CQ. A threshold level of 35% of the EGG signal margin of error in clinical measurement of air pressure vari-
amplitude was used to distinguish between open and closed times ables in phonation, especially during phonation into water.
of the glottis (Figure 5). Figure 6B shows the corresponding airflow resistance curves,
where the differences between the tubes are negligible.
Statistical analyses of the data from subjects The pressure drop in a circular tube conveying incompress-
The measured parameters—CQ, P(sub), P(oral), P(trans), fre- ible viscous steady fluid flow depends on whether the flow is
quency and amplitude of P(oral) oscillation, and SPL—were laminar or turbulent. Laminar flow occurs at lower velocities,
described by calculated mean values and standard deviations. when a fluid flows in parallel layers. Turbulent flow is a less
FIGURE 4. Example of determining bubbling frequency and the peak-to-peak amplitude of the filtered oral pressure signal. Frequency of P(oral)
variation = number of (positive or negative) P(oral) peaks per time. Amplitude of P(oral) variation = mean difference Di between consecutive pos-
itive peak and negative peak = sum(Di)/n, where n is number of pairs of + and − peaks in filtered signal. P(oral), oral pressure.
ARTICLE IN PRESS
4 Journal of Voice, Vol. ■■, No. ■■, 2016
TABLE 1.
Differences for Pressure Values for RT and LVT Within
0–0.3 L/s Flow Rate
P(sub) (cm H2O)
Q (L/s) Glass (2 cm) Lax Vox (2 cm) Diff
0.06 2.96 2.96 0.00
0.10 3.11 3.16 -0.05
0.20 3.42 3.42 0.00
0.30 3.82 3.82 0.00
FIGURE 6. A. Pressure flow curves for Resonance tube and Lax Vox tube when submerged 2 cm and 10 cm in water with three immersion angles.
B. Flow resistance (mean pressure/mean flow) curves for Resonance tube and Lax Vox tube when submerged 2 cm and 10 cm in water with three
immersion angles.
ARTICLE IN PRESS
6 Journal of Voice, Vol. ■■, No. ■■, 2016
FIGURE 7. Pressure flow curves for Resonance tube (glass) submerged in 45° angle and Lax Vox tube submerged in 90° angle in water.
TABLE 2.
Mean Values and Standard Deviations of CQ, Estimated Subglottic, Oral (Supraglottic), and Transglottic Pressure, Fre-
quency and Amplitude of P(oral) Oscillation, and SPL for Phonation into RT and LVT at Two Immersion Depths (2 cm
and 10 cm) and Two Loudness Levels (Habitual and Loud Phonation)
CQ 35% P (sub) P (oral) P (trans) Freq. P (oral) Ampl. P (oral) SPL
Habitual Loudness (%) (cm H2O) (cm H2O) (cm H2O) (Hz) (cm H2O) (dB)
Glass tube 2 cm
Mean 53.8 10.4 4.3 6.1 17.4 1.7 71.7
SD 10.3 3.7 1.4 2.9 1.7 0.7 5
Lax tube 2 cm
Mean 53.8 9.2 4.6 4.6 15.1 2 70.5
SD 12.8 3.1 1.6 2.4 2.5 0.8 2.8
Diff. for tubes (Student / Wilcoxon) ns ns ns p 0,03 p 0,001 p 0,008 ns
Glass tube 10 cm
Mean 55.3 15.7 10.7 7 15.8 1.7 71.6
SD 10.2 3.4 0.9 2.6 2.7 0.7 4.9
Lax tube 10 cm
Mean 54.9 14.6 10.9 5.1 14.6 1.7 71.6
SD 8.1 3 1 2.8 2.4 0.6 3.4
Diff. for tubes (Student / Wilcoxon) ns ns ns p 0,047 p 0,004 ns ns
FIGURE 8. Distribution of (A) CQ, (B) P(trans), and (C) amplitude and (D) frequency of peak-to-peak P(oral) for trained (N = 6) and un-
trained subjects (N = 8) in different tasks (both loudness together). CQ, contact quotient; P(trans), transglottic pressure; P(oral), oral pressure.
very low flow rates (<0.05 L/s) the opposite was measured. We between phonation into the tubes. P(oral) oscillation during water
did not consider flow rates lower than 60 mL/s, as measure- resistance exercising results from water bubbling. Bubbling starts
ment accuracy gets low then, and because very low flow rates as the air pressure inside the tube slightly exceeds the hydro-
are not typically reported in humans. It seems that when im- static pressure in water, and the bubble separates from the tube
mersion depth in water is the same, the differences in length, end when the buoyancy force acting on the bubble is about the
material, and inner diameter of these tubes have rather little effect same or higher than the aerodynamic force in the tube.20 The
on the flow resistance. Some differences were observed in pho- frequency of bubbling is known to increase with flow rate and
nation of the subjects though. with a decrease of tube orifice, and to be practically indepen-
P(trans) was somewhat lower with Lax Vox tube in phona- dent on the depth of orifice immersion if it is between 2.54 cm
tion at habitual loudness, which could imply softer phonation.12 and 25.4 cm.21 Thus, for the same airflow rate, a slightly lower
CQ is typically smaller with softer phonation, but in our results bubbling frequency would be expected for Lax Vox tube com-
there were no significant differences in either CQ, P(sub), or SPL pared with Resonance tube. On average, the frequency of P(oral)
ARTICLE IN PRESS
8 Journal of Voice, Vol. ■■, No. ■■, 2016