Literature Review

Auscultation is a non-invasive and cost-effective physical examination technique that is optimal for prognosis, but interpreting the auditory information of lung breath sounds requires years of experience and specialization [1]. Digital stethoscopes can enhance, record, and store auscultation sounds for future reference, providing a research path for exploring ways to automate diagnosing and monitoring respiratory disease using the sound of breath [2]. Several studies have investigated the use of breath sounds as a potential tool for detecting and monitoring asthma [3][4][5]. Here we investigate the research on the individual contributions of breath sounds from the mouth and chest and how it is effected by asthma.

Keywords

breath sounds lung auscultation mouth breathing chest breathing asthma detection respiratory sounds wheezing airway inflammation asthma diagnosis asthma monitoring sensitivity specificity accuracy adult patients children

Breath sounds from the mouth and chest have been investigated in relation to respiratory diseases, including asthma, in 12 studies. Different methods of recording breath sounds were utilized, with some studies including only a few patients while others included several hundred. Most studies concluded that breath sounds from lung auscultation can provide useful information for diagnosing and monitoring asthma, while few have explored the potential of vocal breath sounds. The reported levels of sensitivity, specificity, and accuracy for the different sources of breath sounds varied across the studies, highlighting the need for standardization and improvement in order for it's utilization in clinical practice.

In Table below, the first column shows the location of the sound from the upper airway to the lower airway with the corresponding sound type in relation to the disease associated with it.

Discussion

Breath sound recordings from various locations have been investigated in multiple studies, which have utilized different study designs, patient populations, sample sizes, and recording methodologies. Despite these variations, all of the studies have concluded that breath sound recordings are an essential biomarker in the determination of underlying respiratory diseases. The studies have employed various clever techniques to classify and determine which features of the signal effectively contain the necessary information. However, due to the differences in study design and methodology, the reported results on the effectiveness of breath sound recordings as a biomarker have been variable. Therefore, standardization of recording methods and data analysis is essential in clinical practice for accurate and consistent use of breath sounds as a diagnostic tool.

1. Analysis of Acoustic Features for Speech Sound Based Classification of Asthmatic and Healthy Subjectslink

1. Recording location: Mouth
2. Data acquired: Sustained phonations of speech sounds: /α:/, /i:/, /u:/, /eI/, /ou/, /s/, and /z/, and non-speech sounds: breath and cough; ZOOM H6 handy recorder was used for all recordings.
3. Dataset: 95 subjects with 47 (28M, 19F) patients and 48 (24M, 24F) healthy.
4. Conclusion: Non-speech stimulus is better than all the speech stimulus. The sound /o℧/ performed the best among all speech stimuli and second best among all speech and non-speech stimuli.

2. Automatic prediction of spirometry readings from cough and wheeze for monitoring of asthma severitylink

1. Recording location: Mouth
2. Data acquired: subjects are asked to cough and wheeze for at least five times, recorded at a sampling rate of 48kHz and 16-bit using the ZOOM H6 handy recorder.
3. Dataset: The recordings used in this study were obtained from a total 28 subjects comprising 16 healthy subjects (10 male and 6 female) and 12 asthmatic patients (7 male and 5 female)
4. Conclusion: The proposed approach predicts FEV1%, FVC% and FEVl_FVC with RMSE of 11.6%, 10.3%, and 0.08 respectively. The 3 asthma severity classification using the predicted spirometry readings results in a classification accuracy of 77.77%.

3. Comparison of Cough, Wheeze and Sustained Phonations for Automatic Classification Between Healthy Subjects and Asthmatic Patientslink

1. Recording location: Mouth
2. Data acquired: Sustained phonations, namely /A:/, /i:/, /u:/, /eI/, /oU/ and compare their classification performances with the cough and wheeze. Recording was done by using ZOOM H6 handy recorder at a sampling rate of 48kHz and 16 bits/sample.
3. Dataset: Each stimuli is recorded for five times in a row. Thus, we obtain 355 recordings (175 for patients and 180 for healthy subjects) for each of the seven stimuli.
4. Conclusion: The experimental results demonstrate that wheeze is the best stimuli for classification with a classification accuracy of 90.5%. However, sustained /ɪː/ performs the best among all sustained vowels with an accuracy of 80.8%. As the best performing stimuli is wheeze where there is no voicing

4. Characteristics of breath sound in infants with risk factors for asthma developmentlink

1. Recording location: Mouth
2. Data acquired: Breath sounds were recorded 10 or more breaths in a silent room using a handheld microphone. The microphone was placed on the right upper anterior chest at the second [intercostal space" />
3. Dataset: A total of 443 infants (mean age, 9.9 months; range, 3-24 months) were included in the present study.
4. Conclusion: The breath sound analysis may be useful for assessing the airways of infants for asthma development.

5. An accurate recording system and its use in breath sounds spectral analysislink

1. Recording location: Trachea
2. Data acquired: Breath sounds are recorded at the trachea simultaneously with the airflow signal at 0.5- and 1-1/s levels.
3. Dataset: Different characteristics of the spectra are calculated in the range 60-1,260 Hz for 11 normal and 10 asthmatic subjects.
4. Conclusion: Breath sounds discriminate between asthmatics and normal subjects in an objective way. It indicates that for each subject frequency spectrum strongly depends on flow rate. Thus a given spectrum cannot be judged “normal” or “abnormal” without reference to the corresponding flow rate. Sound level is higher at the trachea than at any other point of the chest or back.

6. A New Modality Using Breath Sound Analysis to Evaluate the Control Level of Asthmalink

1. Recording location: Trachea + Lungs
2. Data acquired: Breath sounds were recorded using two sensors, located on the right anterior chest and trachea.
3. Dataset: 80 asthmatic children and 59 non-asthmatic children underwent breath sound analysis in an asymptomatic state.
4. Conclusion: Asthma control could be evaluated using a new index calculated from breath sound analysis. The acoustic transfer characteristics between the two points was calculated, which indicated the relationship between frequencies and attenuation during breath sound propagation.

7. Significant differences in flow standardised breath sound spectra in patients with chronic obstructive pulmonary disease, stable asthma, and healthy lungslink

1. Recording location: Trachea + Lungs
2. Data acquired: Breath sounds were recorded simultaneously at the chest and at the trachea
3. Dataset: Flow standardised inspiratory breath sounds in patients with COPD (n = 17) and stable asthma (n = 10) with significant airways obstruction and in control patients without any respiratory disorders (n = 11)
4. Conclusion: The results of this study indicate that the frequency content of breath sounds in asthmatic patients with airways obstruction differs markedly from that of patients with COPD and those with normal lungs.

8. Multichannel lung sound analysis for asthma detectionlink

1. Recording location: Lungs
2. Data acquired: A novel 4-channel data acquisition system from four different positions over the posterior chest, as suggested by the pulmonologist.
3. Dataset: We collected lung sounds of 60 subjects (30 normal and 30 asthma)
4. Conclusion: The proposed multichannel asthma detection method where the presence of wheeze in lung sound is not a necessary requirement, outperforms commonly used lung sound classification methods in this field and provides significant relative improvement.

9. Spectral characteristics of chest wall breath sounds in normal subjects.link

1. Recording location: Lungs
2. Data acquired: Chest wall breath using contact acoustic sensors. Inspiratory and expiratory sounds were picked up at three standard locations on the chest wall during breathing at flows of 1-2 l/s and analysed breath by breath in real time.
3. Dataset: Chest wall breath sounds from 272 men and 81 women were measured using contact acoustic sensors
4. Conclusion: The amplitude spectrum of normal chest wall breath sounds has two linear parts in the log-log plane—low and high frequency segments—that are best characterised by their corresponding regression lines. Four parameters are needed and are sufficient for complete quantitative representation of each of the spectra: the slopes of the two regression lines plus the amplitude and frequency coordinates of their intersection.

10. Measurement and theory of wheezing breath soundslink

1. Recording location: Lungs
2. Data acquired: Characteristics of minimicrophone inserted into the trachea and main bronchi.
3. Dataset: Characteristics of breath sounds in 7 asthmatic and 3 non-asthmatic wheezing patients.
4. Conclusion: The spectral shape, mode of appearance, and frequency range of wheezes with specific predictions of five theories of wheeze production were compared: 1) Turbulence-induced wall resonator, 2) Turbulence-induced Helmholtz resonator, 3) Acoustically stimulated vortex sound (whistle), 4) Vortex-induced wall resonator, and 5) Fluid dynamic flutter. We conclude that the predictions by 4 and 5 match the experimental observations better than the previously suggested mechanisms.

11. A novel method for detecting airway narrowing using breath sound spectrum analysis in childrenlink

1. Recording location: Lungs
2. Data acquired: Breath sounds were recorded for ≥10s using a handheld microphone placed on the right upper anterior chest at the second [intercostal space" />
3. Dataset: A total of 65 children with asthma participated in this study (mean age 9.6 years).
4. Conclusion: The clinical sensitivity of these parameters for bronchial constriction was sufficient, and the breath sound analysis technique is safe and simple to perform. This technique may therefore be useful for pulmonary function testing in infants and young children.

12. Variation of Breath Sound and Airway Caliber Induced by Histamine Challengelink

1. Recording location: Lungs
2. Data acquired: Inspiratory breath sounds were recorded from the chest wall during histamine challenge. The microphone used in this study was held by hand against the posterior chest wall. 4 recordings were made of breath sound during inspiration from the lower right posterior chest wall 3 to 5 em below the inferior angle of the scapula. The design incorporates a Tufnol diaphragm to improve the acoustic match between the chest wall and the electret transducer. The patient was then asked to exhale into a spirometer, and expiratory flow volume curve and FEV.
3. Dataset: 5 patients: 4 male, 1 female
4. Conclusion: Spectral analysis of breath sound may be a useful addition to conventional spirometry in identifying changes in airway diameter.

Conclusion

While the existing studies have primarily focused on utilizing either mouth or chest breath sounds, the combination of both sources may provide a more comprehensive evaluation of respiratory function. Further research is needed to explore the potential of using this approach in clinical practice. Therefore, we hypothesize that breath sounds recorded from mouth and chest simultaneously can yeild better results than when used individually. This needs to be confirmed by experiments by obtaining the necessary data. The findings of these studies could have significant implications for improving asthma diagnosis and monitoring, ultimately leading to better health outcomes for patients.

References

1. Bohadana, Abraham, Gabriel Izbicki, and Steve S. Kraman. "Fundamentals of lung auscultation." New England Journal of Medicine 370.8 (2014): 744-751. link
2. Swarup, Supreeya, and Amgad N. Makaryus. "Digital stethoscope: Technology update." Medical Devices: Evidence and Research (2018): 29-36. link
3. Islam, Md Ariful, et al. "Multichannel lung sound analysis for asthma detection." Computer methods and programs in biomedicine 159 (2018): 111-123. link
4. Bokov, Plamen, et al. "Wheezing recognition algorithm using recordings of respiratory sounds at the mouth in a pediatric population." Computers in biology and medicine 70 (2016): 40-50. link
5. Nagasaka, Yukio. "Lung sounds in bronchial asthma." Allergology International 61.3 (2012): 353-363. link