Indian Journal of Urology
: 2009  |  Volume : 25  |  Issue : 1  |  Page : 92--98

Perineal sound recording for diagnosis of bladder outlet obstruction

Tim Idzenga1, Johan J.M Pel2, Ron van Mastrigt1,  
1 Department of Urology, Sector Furore, Erasmus MC, Rotterdam, Netherlands
2 Department of Neuroscience, Vestibular Group, Erasmus MC, Rotterdam, Netherlands

Correspondence Address:
Tim Idzenga
Deptartment of Urology, Erasmus MC, Room Ee1630, PO Box 2040, 3000 CA Rotterdam


Objectives: Elderly men are prone to developing lower urinary tract symptoms (LUTS) possibly caused by bladder outlet obstruction (BOO). The most frequently used method to diagnose this condition is an invasive and time-consuming pressure-flow study. We are developing a novel non-invasive method to diagnose BOO in men with LUTS based on perineal sound recording. Methods: A biophysical model urethra was made from polyvinyl alcohol (PVA) cryogel with viscoelastic properties comparable to those of the male pig urethra. To this model different degrees of obstruction were applied and sound was recorded at different positions downstream of the obstruction. In a study in 16 healthy male volunteers the variability and repeatability of perineal sound recording was tested. Results: In the model three parameters, derived from the frequency spectrum of the recorded sound (i.e., weighted average frequency, standard deviation and skewness) are uniquely related to the degree of obstruction (linear regression, P<0.001). The variability of perineal sound recording in healthy male volunteers was found to be smaller within volunteers than between volunteers (Kruskal-Wallis, P<0.001) and the repeatability was comparable to that of the maximum flow rate. Conclusions: We conclude that perineal sound recordings are significantly different between volunteers. In combination with the unique relations found in the model-experiments these results increase the probability that perineal sound recording can be used as a simple and cheap non-invasive method to diagnose BOO. Clinical testing of this method is therefore strongly indicated.

How to cite this article:
Idzenga T, Pel JJ, Mastrigt Rv. Perineal sound recording for diagnosis of bladder outlet obstruction.Indian J Urol 2009;25:92-98

How to cite this URL:
Idzenga T, Pel JJ, Mastrigt Rv. Perineal sound recording for diagnosis of bladder outlet obstruction. Indian J Urol [serial online] 2009 [cited 2021 Jan 26 ];25:92-98
Available from:

Full Text


Elderly men are prone to developing lower urinary tract symptoms (LUTS), such as a weak urinary stream, frequent (nocturnal) voiding and incomplete emptying of the bladder. The two most probable causes for these symptoms are a weakly contracting bladder muscle or increased urethral resistance (bladder outlet obstruction; BOO). In most cases, the voiding symptoms are caused by BOO, which is, in turn, frequently caused by benign prostatic enlargement (BPE). The most frequently used method to diagnose this condition in men with voiding symptoms is an invasive pressure-flow study. [1] This testing method however is time-consuming, expensive, and patient-unfriendly. Therefore it is often decided to perform a transurethral resection of the prostate (TURP) based on only a flow rate measurement. A low flow rate can also indicate a weakly contracting bladder and in that case, TURP does not improve the symptoms. To make urodynamic testing of patients with LUTS easier several alternative non-invasive methods have been developed. [2],[3],[4],[5] Some of these methods require interruption/manipulation of the voiding [2],[3],[4] or expensive equipment. [5] We are developing a novel non-invasive method to diagnose BOO. This method is based on sound, recorded during voiding, using a simple microphone pressed against the perineum [Figure 1]. It is hypothesized that this recorded sound is related to the degree of obstruction.

 Methods (1)


The hypothesis that sound, recorded during voiding, could be used as a diagnostic method originated in the mid-sixties with a voiding audiograph. [6] It was thought that the splashing sound of urine voided in a cup, recorded using a microphone and corrected for the voided volume, was a measure for bladder pressure. One decade later, the microphone was moved from the cup to the body at the level of the perineum and the sound produced in the urethra was recorded simultaneously with the flow rate. [7] It was hypothesized that urinary flow is turbulent at the bladder neck. These turbulences cause pressure fluctuations on the wall of the urethra that can be recorded as sound transmitted via the urethra to the skin. [8] In silicone [9] or latex [10] test tubes it was shown that pressure fluctuations are indeed induced by flow and that the power spectrum of the recorded sound depended among others on the degree of obstruction applied. However, it was not known how the degree of obstruction could be derived from the power spectrum.

In a review on bio-acoustic signals it was noted that phonoangiographic methods, developed in cardiology, might be non-invasive and inexpensive, but that clinical application of these methods has been overtaken by large-scale application of ultrasound Doppler. [11] Nevertheless, a method based on perineal sound recording would make diagnosis of BOO cheap and simple. Therefore, over the years attempts have been made to relate sound recordings from biological fluids (e.g., blood, urine) to the obstruction passed (e.g., stenosis, prostatic obstruction). Despite encouraging results, a non-invasive measurement device is not yet available for urology. One limiting factor is that the test tubes used in earlier experiments are rather stiff compared to the more flexible and distensible blood vessel or urethra, so that results obtained in these test tubes cannot be translated to the clinical situation. Recently, it was reported that flexible and distensible tubes can be made from a 10% aqueous solution of polyvinyl alcohol (PVA) cryogel. [12] The number of freeze-thaw cycles, the rate of freezing/thawing and the concentration of the PVA control the viscoelastic properties of this gel.

To test the hypothesis that sound recorded downstream of an obstruction is related to the degree of the obstruction, we constructed PVA models with varying viscoelastic properties. [13] To these models different degrees of obstruction were applied and the position of the recording microphone relative to the obstruction was varied. The average amplitude (A) of the recorded sound and the weighted average frequency (ƒ c ) of its frequency spectrum were found to depend significantly on the degree of obstruction.

Results (1)

Both parameters appeared to increase with increasing degree of obstruction. This result supported the hypothesis that a non-invasive diagnostic method for BOO might be based on perineal sound recording. Both and depended significantly on the microphone position and on the viscoelastic properties of the PVA-model. This made it necessary to develop a model of the urethra with viscoelastic properties as close as possible to those of the human male urethra.

 Methods (2)

A biophysical model of the male urethra

To establish which model urethra best represents the human male urethra the viscoelastic properties of different models were measured. The number of freeze-thaw cycles of the cryogel, the profile of the flow channel in the model and the number of cross-sectional layers were varied. The rate of freezing/thawing and the concentration of the PVA cryogel were kept constant in all models. To select the model urethra with viscoelastic properties comparable to those of the human male urethra the male pig urethra was used as a reference because of the similarities between pig and human physiology. We measured the viscoelastic properties using the method developed by Coolsaet et al. [14] and adapted it for use in cylindrical structures. This method involves the stepwise application of strain to the urethral wall by rapidly injecting known volumes of water into the sealed urethra and recording the pressure response.

12 different model urethras were constructed by pouring a 10% aqueous solution of PVA into a cylindrical mould. After 6 h of rest at room temperature (21°C), the mould was stored in a freezer at -20°C. It remained in the freezer for 14 h and was then stored at room temperature again for 10 h. This completed one freeze-thaw cycle. To create a channel allowing flow through the models, either a strip with a Y-profile (with legs 5 mm wide, [Figure 2a]) or a circular rod (diameter 4 mm) was placed along the central axis of the mould. The Y-profile reflected the lumen of a male pig urethra without flow [Figure 2b] and the circular profile reflected the expected profile of a male pig urethra during flow. From 8 male pigs sacrificed at the Department of Experimental Cardiology (Erasmus MC, Rotterdam, The Netherlands) bladder/urethra preparations were made. The part of the urethra closest to the bladder (proximal part) and distal to the prostate was selected. The average (±1 SD) length of the sections was 65 × 16 mm. An example of a cross-section is shown in [Figure 2b].

The models were placed in a water-filled container (approximately 1 cm below the water-surface) at room temperature and the pig urethras were placed in the same container (approximately 1 cm below the solution-surface) but filled with a cold (10°C) modified Krebs solution (118 mM NaCl; 4.7 mM KCl; 25 mM NaHCO 3 ; 1.2 mM KH 2 PO 4 ; 1.8 mM CaCl 2 ; 1.2 mM MgSO 4 , and 11 mM glucose aerated with 95% O 2 and 5% CO 2 at pH 7.4). One side of the model or pig urethra was connected to a 5 ml syringe to inject a known volume of water in a very short time. The pressure response was recorded using a disposable pressure transducer connected to the other side of the urethra. Increasing volumes were injected, starting at 0.2 ml with increments of 0.2 ml to a maximum of 5 ml, or to the volume at which the range of the pressure transducer (~250 cm H 2 O) was exceeded. After each measurement the injected water was extracted. The transducer signal was sampled at a frequency of 1000 Hz using an AD-converter (PCL-818, Advantech® ) and stored in a PC for further analysis using custom-written programs in Matlab® . For an approximate calculation of the viscoelastic properties of the model and pig urethras from the recorded pressure response a mechanical model of the urethral wall was constructed, consisting of springs, dashpots, and a mass. [15] The simplified analytical solution for the step-response of this model was fitted to 1000 samples (equivalent of 1 s) of the pressure signal.

Results (2)

One of the results found in the model urethras was that stress increased more rapidly as a function of strain with increasing number of freeze-thaw cycles. This implies increased stiffness, which was also found by Chu and Rutt. [12] Furthermore, two time-constants (τ1,2 ) were found for the stress-relaxation in the model and pig urethral wall. In the male pig urethra, these two time-constants differed by a factor of ~10 from each other, which supports the assumption that two independent relaxation processes are activated after stepwise application of strain. Statistical analysis (Mann-Whitney U-test) was used to compare the viscoelastic properties measured in model urethras to those measured in male pig urethras. Based on this comparison we concluded that a model urethra with a Y-shaped flow channel that was uniformly freeze-thawed three times showed best agreement with the male pig urethra. Therefore, this model was selected to study the relation between recorded sound and the degree of obstruction.

 Methods (3)

Recorded sound related to the degree of obstruction?

The flexible and extensible biophysical model which we developed [15] was placed in a container filled with water to prevent evaporation. Different degrees of obstruction were imposed by inflating a blood pressure cuff around the model and sound was recorded at different positions downstream of the obstruction. [16] One side of the model was connected to a water column with regulated water level and the other side to an outflow tube, which drained into a rotating-disk flow meter. The blood pressure cuff formed a distensible obstruction of the model urethra, similar to an elastic prostate, and was connected to a second water column. The level in the first water column (bladder pressure) was varied between 60 and 180 cm and in the second column (cuff) between 20 and 180 cm (both in steps of 20 cm) resulting in 63 possible combinations. Only in 35 combinations of the two column levels a flow rate higher than 2 ml/s was realized. The degree of obstruction was quantified by the bladder outlet obstruction index (BOOI): [17]

BOOI = P det -2.Q max , (1)

with P det (cmH 2 O) being the detrusor pressure at maximum urinary flow rate, modeled by the pressure-drop over the model urethra, and Q max (ml/s) the maximum flow rate through it. According to the international golden standard [1] an unobstructed urethra is defined by BOOI 40. For 20 c (Hz), the standard deviation ? (Hz), and the skewness γ1 (-). The parameters derived from the recorded sound might depend on: the bladder pressure, the flow rate, and the cuff pressure. In the model urethra these three variables are related, any one of the three can be calculated from the other two. Therefore each parameter can be described as a function of two independent variables, e.g., bladder pressure and flow rate.

Results (3)

All three parameters were uniquely related to the degree of obstruction (Linear Regression coefficient: -0.568, -0.653 and 0.466 respectively, P c and σ decreased with increasing degree of obstruction and γ1 increased. For all three parameters the mean was found to be significantly different between recordings from obstructed and unobstructed models (Student's t-test, P c means a shift of the center of gravity of the frequency spectrum to lower frequencies and a decrease in σ means that the power is more concentrated around this center of gravity. Both indicate that increasing the degree of obstruction adds relatively more power to the lower section of the frequency spectrum, caused by the turbulence in the urinary flow.

The results found in this study are based on experimental values in a biophysical model of the urethra. In humans, anatomical aspects of the obstruction and the lower urinary tract may vary between individuals and possibly affect the sound recording. Therefore, further testing of perineal sound recording in male volunteers and patients is needed to establish the diagnostic value of this method.

 Methods (4)

Perineal sound recording in male volunteers

Based on the findings in our model-experiments a measurement setup for perineal sound recording in healthy male volunteers and patients was developed. 16 volunteers were recruited to void at least ten times at normal desire. [18] They filled out an IPSS-form as a measure of LUTS and voided in a standing position with a microphone pressed to the perineum directly behind the scrotum using a modified jockstrap. The microphone consisted of a modified dual head stethoscope (MDF Instruments Europe, Copenhagen, Denmark) with the earpiece replaced by a piezoceramic transducer. The voided volume was measured with an external weight transducer. Differentiation of the voided volume with respect to time resulted in the urinary flow rate.

From each sound recording, the part with a flow rate higher than 50 percent of the maximum was selected [see [Figure 4]]. For this section, the frequency spectrum was calculated between 100 and 1000 Hz in a time-window of ~0.8 s using fast Fourier transform. The time-window was than shifted ~0.2 s and the calculation was repeated. This procedure was repeated until the end of the recorded signal was reached. Each frequency spectrum was characterized by the three parameters found in the model-experiments. [16] For each recording subsequently the median value of each parameter was calculated. Also the maximum of the flow rate signal (Q max ) was assessed.

Results (4)

The variability of all parameters was significantly smaller within volunteers than that between volunteers (Kruskal-Wallis, P [19] For each pair of recordings we calculated the difference and the mean of the values of the three parameters and the maximum flow rates. The normalized statistic was calculated as the standard deviation of the differences, divided by the range of the means. The repeatability measures of the three parameters were not significantly different from that of the maximum flow rate (Wilcoxon's signed rank test, P=0.438, P=0.163 and P=0.756, respectively). The variation between the volunteers is most likely a result of physiological variation, e.g. differences in structure and function of the lower urinary tract. The sound recordings in this study are similar to the examples presented by Koiso et al. [20] and Teriφ. [9] The values for the weighted average frequency given by Teriφ for the one volunteer in his study (112 and 129 Hz) are comparable to the values we found in our volunteer population (125-240 Hz). However, the values given for the patients (170-454 Hz) are higher than those in our population. Even though the values are calculated from a slightly different part of the sound frequency spectrum this hints at a possible way to discriminate healthy volunteers from patients.

We calculated the correlation between parameters using the non-parametric Spearman's correlation coefficient. The weighted average frequency had a good correlation with the standard deviation (Spearman's rho = 0.720, P [16] [Figure 4].


Based on these findings we conclude that perineal sound recording gives repeatable results in healthy male volunteers and that the results are significantly different from volunteers. In combination with the unique relation between recorded sound and the degree of obstruction we found in the model-experiments, these results increase the probability that perineal sound recording can be used as a very simple and cheap method to non-invasively diagnose BOO. Clinical testing of this method is therefore strongly indicated.


This research is supported by the Dutch Technology Foundation STW, applied science division of NWO and the Technology Program of the Ministry of Economic Affairs. Parts of the text have been reproduced [16] with kind permission from IEEE and [13],[15],[18] with kind permission from John Wiley and Sons, Inc.


1Griffiths DJ, Hφfner K, van Mastrigt R, Rollema HJ, Spεngberg A, Gleason DM. Standardization of terminology of lower urinary tract: Pressure-flow studies of voiding, urethral resistance and urethral obstruction. Neurourol Urodyn 1997;16:1-18.
2Pel JJ, van Mastrigt R. The variable outflow resistance catheter: A new method to measure bladder pressure noninvasively. J Urol 2001;165:647-52.
3Sullivan MP, Yalla VS. Penile urethral compression-release maneuver as a non-invasive screening test for diagnosing prostatic obstruction. Neurourol Urodyn 2000;19:657-69.
4McRae LP, Bottacini MR, Gleason DM. Noninvasive quantitative method for measuring isovolumetric bladder pressure and urethral resistance in the male I: Experimental validation of the theory. Neurourol Urodyn 1995;14:101-14.
5Ding YY, Ozawa H, Yokoyama T, Nasu Y, Chancellor MB, Kumon H. Reliability of color doppler ultrasound urodynamics in the evaluation of bladder outlet obstruction. Urology 2000;56:967-71.
6Keitzer WA, Huffman GC. The voiding audiograph: A new voiding test. J Urol 1966;96:404-10.
7Bradley WE, Brockway BP, Timm GW. Auscultation of urinary flow. J Urol 1977;118:73-5.
8van Koeveringe AJ, van Mastrigt R. A relation between the sound produced by urethral turbulence in patients and objectively assessed subvesical obstruction. Neurourol Urodyn 1991;10:442-3.
9Teriφ H. Acoustic method for assessment of urethral obstruction: A model study. Med Biol Eng Comput 1991;29:450-6.
10Jones SA, Fronek A. Analysis of break frequency downstream of a constriction in a cylindrical tube. J Biomech 1987;20:319-27.
11Ask P, Hφk B, Loyd D, Teriφ H. Bio-acoustic signals from stenotic tube flow: State of the art and perpsectives for future methodological development. Med Biol Eng Comput 1995;33:669-75.
12Chu KC, Rutt BK. Polyvinyl Alcohol cryogel: An ideal phantom material for MR studies of flow and elasticity. Magn Reson Med 1997;37:314-9.
13Idzenga T, Pel JJ, Baldewsing RA, Mastrigt R. Perineal noise recording as a non-invasive diagnostic method of urinary bladder outlet obstruction: a study in polyvinyl alcohol and silicone model urethras. Neurourol Urodyn 2005;24:381-8.
14Coolsaet BL, van Duyl WA, van Mastrigt R, van der Zwart A. Step-wise cystometry of urinary bladder: New dynamic procedure to investigate viscoelastic behavior. Urology 1973;2:255-7.
15Idzenga T, Pel JJ, van Mastrigt R. A biophysical model of the male urethra: Comparing viscoelastic properties of polyvinyl alcohol urethras to male pig urethras. Neurourol Urodyn 2006;25:451-60.
16Idzenga T, Pel JJ, van Mastrigt R. Towards an acoustic non-invasive diagnosis of urinary bladder outlet obstruction. IEEE Trans Biomed Engg 2008;55:1764-71.
17Abrams P. Bladder outlet obstruction index, bladder contractility index and bladder voiding efficiency: Three simple indices to define bladder voiding function. BJU Int 1999;84:14-5.
18Idzenga T. Variability and repeatability of perineal noise recording in a population of healthy male volunteers. Neurourol Urodyn 2008; Jun 12. [Epub ahead of print].
19Mastrigt R, Huang Foen Chung JW. Comparison of repeatability of non-invasive and invasive urodynamics. Neurourol Urodyn 2004;23:317-21.
20Koiso K, Nemoto R, Ohtani M. Urographic studies of benign prostatic hypertrophy. J Urol 1991;145:1071-7.