Introduction
Nasal airflow measurements such as rhinomanometry consist of passive, active and acoustic rhinometry, which is based on the analysis of a sound pulse generated within the nasal passages. Acoustic rhinometry can calculate and display the minimal cross-sectional areas (MCA) of the nose at various distances from the nostril, and the total volume of the nose.
For many years, multiple new methods (in addition to rhinomanometry and acoustic rhinometry) have attempted to measure nasal obstruction objectively, such as impulse oscillometry,Reference Van Erck, Votion, Art and Lekeux1 long-term rhinoflowmetry,Reference Grutzenmacher, Lang, Mlynski, Mlynski and Mlynski2 optical rhinometryReference Wustenberg, Huttenbrink, Hauswald, Hampel and Schleicher3 and nasal spirometry.Reference Cuddihy and Eccles4 One recent development is a new software program, Odiosoft-Rhino, which automatically analyses the spectral parameters of nasal sound via a new, versatile, personal computer based software package.Reference Seren5
In this study, we investigated the use of the Odiosoft-Rhino program (Odiosoft Medical Software, Giresun, Turkey) to analyse the nasal sounds of patients with nasal obstruction due to known nasal septal deviation. For each patient, we compared Odiosoft-Rhino results with those for acoustic rhinometry and nasal symptom scoring.
Materials and methods
We performed Odiosoft-Rhino testing and acoustic rhinometry on 61 patients with known nasal septal deviations. None of the patients had a history of nasal trauma or surgery, and none were using nasal medications that could modify nasal function. All patients were examined using nasal endoscopy. We selected patients randomly from the archives of examination cards, and asked each to volunteer for the study. We chose patients with unilateral nasal septal deviation without concha hypertrophy, and only with same side, unilateral nasal obstruction symptoms. The primary complaint of the patients was nasal obstruction, and none demonstrated any additional pathology. All patients tested negative with the ALK-Abello allergy test panel. Visual analogue scores for nasal obstruction (on a scale of 0–10 mm) were self-noted for one month. This study was carried out with the permission of the medical research ethics committee of Istanbul University and with the informed consent of the subjects.
Acoustic rhinometry measurements were completed using RhinoScan software (4.2.0.0, version 3.01.Xf/v 3.0b; Interacoustics AS, Assens, Denmark). Minimal cross-sectional areas at 2.2 and 5.5 cm from the nostril were noted for both nasal cavities. Three to four measurements were carried out and the mean value recorded, unless the difference between the measurements was greater than 10 per cent. When the difference was greater than 10 per cent, the test was re-performed until the difference became less than 10 per cent.
Nasal expiration sound spectral analysis at 200–500 Hz, 500–1000 Hz, 1000–2000 Hz, 2000–4000 Hz and 4000–6000 Hz intervals was undertaken for both nasal cavities. We compared patients' Odiosoft-Rhino data, acoustic rhinometry results and visual analogue scores for both cavities separately. The sound of the nasal expiration airflow was recorded with a microphone, which was low-pass filtered at 10 kHz, high-pass filtered at 150 Hz and amplified by 20 dB. The microphone was connected to the sound card and recorded the nasal sounds as a digital sample, with the following specifications: 1–16 bit sampling, 2–44 100 Hz, and three-mono. The Odiosoft-Rhino software was programmed with object-oriented programming (Visual Basic 6.0, Microsoft, Seattle, Washington, USA). This software analysed the nasal sound using the fast Fourier transform and then displayed the input and output waveforms and spectra, as well calculating the magnitude and phase of the transfer function. We had previously calculated the intensity and frequency of nasal sound using this program.
According to our test protocol, the patients were taken to the examination room 20 minutes before the test, seated and asked to wait. All crusts and mucous in the nasal cavity were removed prior to testing. The room was quiet (between 200 and 1000 Hz and under 10 dBL(loudness) sound intensity), with a temperature between 22° and 25°C and a humidity ratio of 50–60 per cent. All artefact sounds (e.g. computer fan noise) were removed with the help of spectrogram overview fast Fourier transform. Additionally, we asked patients to expire from the nose with some force, in order to examine whether this could cause a spectral frequency shift. As a standard procedure during the experiment, four consecutive nasal expirations were performed, and the sound of the expiration airflow was recorded with a microphone which was connected to the sound card. The expiration with the best sound sample rate (44 100 bit/second) was chosen for evaluation. While testing the right side of the nose, the thumb of the left hand gently closed the left nostril, and vice versa. During the test, the probe in the right hand was held parallel to the right nasal cavity. The nasal probe was not inserted into the nasal cavity, but was placed 1 cm away from the nose, using a device which inserted into the microphone and could then be angled toward the jaw (Figure 1). Recording commenced during non-forced nasal expirations. Testing was similarly completed for the left side. Topical nasal decongestants were not used.
Fig. 1 (a) Use of Odiosoft-Rhino equipment, and (b) a sample of graphic result, at varying frequency intervals, in dB.
A p value of less than 0.05 was considered significant. Student t-tests and Pearson correlations were used in the statistical analysis.
Fig. 2 Correlation between visual analogue score (VAS) for right-sided nasal obstruction, and right-sided Odiosoft-Rhino findings, at the 2000–4000 Hz interval.
Results
Acoustic rhinometry and Odiosoft-Rhino testing were performed on 61 patients with known nasal septal deviations. Thirty-seven patients were female (60.65 per cent) and 24 were male (39.35 per cent). Their ages ranged between 18 and 66 years of age (mean age 32.10 ± 11.05 years). Patients' visual analogue scores of their nasal obstruction ranged between 5 and 9 cm (mean score 6.88 ± 2.65 cm). Twenty-five patients (41 per cent) had right-sided septal deviations and 36 (59 per cent) had left-sided deviations. Table I shows the nasal sound spectral analysis for left and right nasal cavities, and the minimal cross-sectional areas 2.2 and 5.5 cm from the nostril, for the left and right nasal cavities.
Table I Patients' odiosoft-rhino and acoustic rhinometry findings
Data are expressed as mean values ± standard deviation. MCA1 = minimal cross-sectional area 2.2 cm from the nostril; MCA2 = minimal cross-sectional area 5.5 cm from the nostril
No significant difference was detected between the findings for men and women (p > 0.05). All data showed a normal distribution when expressed as the mean value plus the standard deviation. Results for the 2000–4000 Hz and 4000–6000 Hz interval findings (dB) for the deviated side of the nose were found to be significantly higher than those for the other side of the nose. The minimal cross-sectional area 2.2 cm from the nostril on the deviated side of the nose was found to be significantly smaller than that on the other side. The other interval findings of Odiosoft-Rhino testing were not statistically significant (p > 0.05) (Table II). We found a weak correlation between the Odiosoft-Rhino results and the minimal cross-sectional area 2.2 cm from the nostril, for the right side in patients with right-sided deviations, as follows: at 2000–4000 Hz, correlation coefficient r = 0.493, p < 0.01; and at 4000–6000 Hz, r = 0.501, p < 0.01. We also found weak correlations between the Odiosoft-Rhino results and the minimal cross-sectional area 2.2 cm from the nostril, for the left side in patients with left-sided deviations, as follows: at 2000–4000 Hz, r = 0.485, p < 0.01; and at 4000–6000 Hz, r = 0.368, p < 0.01) (Table II). Additionally, there was a correlation between the visual analogue scores of nasal obstruction and the Odiosoft-Rhino findings for the right side, in right-sided deviation patients: at 2000–4000 Hz, r = 0.41, p < 0.01 (Figure 2). Similarly, there was a correlation between the visual analogue scores of nasal obstruction and the Odiosoft-Rhino findings for the left side, in left-sided deviation patients: at 2000–4000 Hz, r = 0.30, p < 0.05) (Figure 3).
Table II Comparison and correlation of odiosoft-rhino* and acoustic rhinometry† findings, between deviated and non-deviated sides of nose
* At 2000–4000 Hz and 4000–6000 Hz frequency intervals; †MCA1 values. AR = acoustic rhinometry; **p<0.01. MCA1=minimal cross-sectional area 2.2 cm from the nostril
Seven patients had a severe deflection of the nasal septum which caused near-total obstruction of the nasal passage. In these patients, the 2000–4000 Hz interval Odiosoft-Rhino findings were above 35 dB, and there was a correlation between the severity of deflection and the increase of intensity of the nasal sound in the 2000–4000 Hz interval (r = 0.504, p < 0.01).
Fig. 3 Correlation between visual analogue score (VAS) for left-sided nasal obstruction, and left-sided Odiosoft-Rhino findings, at the 2000–4000 Hz frequency interval.
Discussion
Computer technology has provided new insights into acoustic mechanisms and new measurements of clinical relevance. The joining of pulmonary acoustics with traditionally measured lung mechanics (e.g. air flow and volume), combined with the use of digital techniques to extract information from average sounds under standardised conditions, have all been major steps which have advanced the utility of lung sounds well beyond the realms of the stethoscope. Respiratory acoustic measurements have also been promising in the investigation of upper airway pathologies, for example, in patients with obstructive sleep apnoea or with tracheal narrowing.Reference Pasterkamp, Kraman and Wodıcka6 Some researchers have investigated digitalising respiratory sound.Reference Cheetham, Giordano and Helisto7–Reference Charbonneau, Ademovic, Cheetham, Malmberg, Vanderschoot and Sovijärvi10
Seren has developed a new software program which analyses nasal sound, and has demonstrated that the frequency of nasal sound is an indicator of the intensity of nasal airflow.Reference Seren5 Tahamiler et al. have suggested that the Odiosoft-Rhino software program might be a useful method for evaluating nasal airflow in normal subjects and in patients with allergic rhinitis.Reference Tahamiler, Edizer and Canakcioglu11, Reference Tahamiler, Edizer, Canakcioglu, Guvenc, Inci and Dirican12 In this study, we attempted to evaluate the reliability and validity of Odiosoft-Rhino testing in deviated noses. However, further studies with larger patient populations are required to determine the validity of Odiosoft-Rhino testing.
Rapid expiration affects the intensity of sound only in the related frequency interval and does not cause a shift in frequency spectra to higher intervals. For example, the frequency of a musical note such as C# will be the same, whether performed in a fast or slow manner. In this example, the variable is intensity; however, in the case of turbulence of nasal air, in addition to intensity change, the main variable is the shift of sound from low to high frequency intervals.
We investigated the use of Odiosoft-Rhino software to analyse the nasal sounds of patients with nasal obstruction due to nasal septal deviation. We chose patients who had unilateral nasal septal deviation, without concha hypertrophy or allergic rhinitis, and always with same side unilateral nasal obstructing symptoms.
The only difficulty with Odiosoft-Rhino testing is that it must be performed in a quiet room, because extraneous noise can affect the results. However, acoustic rhinometry tests are also affected by noise. The advantages of the Odiosoft-Rhino method are that it is easy, noninvasive, very rapid (it takes only a minute to perform), requires less cooperation with the subjects than acoustic rhinometry, produces results that are not changed by nasal cannulation (as in acoustic rhinometry), and produces nasal expiratory sounds that can be saved as digital data. The test was not affected by gender or age. In addition, with continuing development, analysis and recording could become faster. Our results showed that the expiratory nasal sound was significantly louder on the obstructed side, at both the 2000–4000 Hz and 4000–6000 Hz intervals. It seems that the findings from these intervals are more sensitive, and therefore more useful than those from any other sound intensity intervals. The potential impact of the nasal cycle will be equal in both acoustic rhinometry and Odiosoft-Rhino testing, because both techniques were carried out during the same session, all within the same period of time.
Mamikoglu et al. showed that acoustic rhinometry could differentiate nasal cavities with and without septal deviation.Reference Mamikoglu, Houser, Akbar, Ng and Corey13 We found statistically significant differences between the minimal cross-sectional area 2.2 cm from the nostril for both left- and right-deviated nasal cavities, and the Odiosoft-Rhino and acoustic rhinometry results showed correlation in patients with septal deviation. Thus, our results demonstrate that the Odiosoft-Rhino technique can be used to differentiate the normal nasal cavity from the obstructed nasal cavity.
• This study aimed to investigate use of a new software program, Odiosoft-Rhino, to determine nasal obstruction by analysing the sounds of nasal expiration
• Odiosoft-Rhino testing was noninvasive, required minimal cooperation and experience, and provided results that could be saved as digital data
• Odiosoft-Rhino findings were strongly correlated with acoustic rhinometry results and visual analogue symptom scores of nasal obstruction
At present, Odiosoft-Rhino testing is not suitable for the determination of size and location of minimal cross-sectional area. However, the future development of intranasal nano-microphones which can be passed from the nasopharynx to the nostrils may allow detection of the site of obstruction. A greater number of studies, and some technical modifications, will be required. Follow-up programmes for operative intervention and long-term medical treatment are time-consuming procedures. We speculate that physicians may be able to better manage their patients by using Odiosoft-Rhino software – patients could record their nasal sounds as digital sound files and send these via the internet to their physician for immediate evaluation.Reference Seren14
Conclusion
The Odiosoft-Rhino test is noninvasive, requires little patient cooperation and minimal physician experience, and produces results that can be saved as digital data. In addition, we have shown that our Odiosoft-Rhino results correlated with both acoustic rhinometry results and visual analogue scores of nasal obstruction. Sound intensity results from the 2000–4000 Hz and 4000–6000 Hz intervals were more sensitive and therefore more useful than those from other sound intensity intervals. Thus, we speculate that Odiosoft-Rhino testing could be used as a novel diagnostic method to evaluate nasal airflow in clinical practice. However, further studies with larger patient populations are required to determine the validity of Odiosoft-Rhino testing.
