Classification of asthmatic breath sounds: preliminary results of the classifying capacity of human examiners versus artificial neural networks

Comparative study of respiratory sounds classification methods based on cepstral analysis and artificial neural networks

In this paper, we investigated and evaluated various machine learning-based approaches for automatically detecting wheezing sounds. We conducted a comprehensive comparison of these proposed systems, assessing their classification performance through metrics such as Sensitivity, Specificity, and Accuracy. The main approach to developing a machine learning-based system for classifying respiratory sounds involved the combination of a technique for extracting features from an unknown input sound with a classification method to determine its belonging class. The characterization techniques used in this study are based on the cepstral analysis, which was extensively employed in the automatic speech recognition field. While MFCC (Mel-Frequency Cepstral Coefficients) feature extraction methods are commonly used in respiratory sounds classification, our study introduces a novelty by employing GFCC (Gammatone-Frequency Cepstral Coefficients) and BFCC (Bark-Frequency Cepstral Coefficients) for this purpose. For the classification task, we employed two types of neural networks: the MLP (Multilayer Perceptron), a feedforward neural network, and a variant of the LSTM (Long Short-Term Memory) recurrent neural network called BiLSTM (Bidirectional LSTM). The proposed classification systems are evaluated using a database consisting of 497 wheezing segments and 915 normal respiratory segments, which are recorded from individuals diagnosticated with asthma and individuals without any respiratory issues, respectively. The highest classification performance was achieved by the BFCC-BiLSTM model, which demonstrated an exceptional accuracy rate of 99.8%.

Digital Pulmonology Practice with Phonopulmography Leveraging Artificial Intelligence: Future Perspectives Using Dual Microwave Acoustic Sensing and Imaging

Respiratory disorders, being one of the leading causes of disability worldwide, account for constant evolution in management technologies, resulting in the incorporation of artificial intelligence (AI) in the recording and analysis of lung sounds to aid diagnosis in clinical pulmonology practice. Although lung sound auscultation is a common clinical practice, its use in diagnosis is limited due to its high variability and subjectivity. We review the origin of lung sounds, various auscultation and processing methods over the years and their clinical applications to understand the potential for a lung sound auscultation and analysis device. Respiratory sounds result from the intra-pulmonary collision of molecules contained in the air, leading to turbulent flow and subsequent sound production. These sounds have been recorded via an electronic stethoscope and analyzed using back-propagation neural networks, wavelet transform models, Gaussian mixture models and recently with machine learning and deep learning models with possible use in asthma, COVID-19, asbestosis and interstitial lung disease. The purpose of this review was to summarize lung sound physiology, recording technologies and diagnostics methods using AI for digital pulmonology practice. Future research and development in recording and analyzing respiratory sounds in real time could revolutionize clinical practice for both the patients and the healthcare personnel.

A temporal dependency feature in lower dimension for lung sound signal classification

Respiratory sounds are expressed as nonlinear and nonstationary signals, whose unpredictability makes it difficult to extract significant features for classification. Static cepstral coefficients such as Mel-frequency cepstral coefficients (MFCCs), have been used for classification of lung sound signals. However, they are modeled in high-dimensional hyperspectral space, and also lose temporal dependency information. Therefore, we propose shifted \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\delta $$\end{document}δ-cepstral coefficients in lower-subspace (SDC-L) as a novel feature for lung sound classification. It preserves temporal dependency information of multiple frames nearby same to original SDC, and improves feature extraction by reducing the hyperspectral dimension. We modified EMD algorithm by adding a stopping rule to objectively select a finite number of intrinsic mode functions (IMFs). The performances of SDC-L were evaluated with three machine learning techniques (support vector machine (SVM), k-nearest neighbor (k-NN) and random forest (RF)) and two deep learning algorithms (multilayer perceptron (MLP) and convolutional neural network (cNN)) and one hybrid deep learning algorithm combining cNN with long short term memory (LSTM) in terms of accuracy, precision, recall and F1-score. We found that the first 2 IMFs were enough to construct our feature. SVM, MLP and a hybrid deep learning algorithm (cNN plus LSTM) outperformed with SDC-L, and the other classifiers achieved equivalent results with all features. Our findings show that SDC-L is a promising feature for the classification of lung sound signals.

Wheeze sound analysis using computer-based techniques: a systematic review

Wheezes are high pitched continuous respiratory acoustic sounds which are produced as a result of airway obstruction. Computer-based analyses of wheeze signals have been extensively used for parametric analysis, spectral analysis, identification of airway obstruction, feature extraction and diseases or pathology classification. While this area is currently an active field of research, the available literature has not yet been reviewed. This systematic review identified articles describing wheeze analyses using computer-based techniques on the SCOPUS, IEEE Xplore, ACM, PubMed and Springer and Elsevier electronic databases. After a set of selection criteria was applied, 41 articles were selected for detailed analysis. The findings reveal that 1) computerized wheeze analysis can be used for the identification of disease severity level or pathology, 2) further research is required to achieve acceptable rates of identification on the degree of airway obstruction with normal breathing, 3) analysis using combinations of features and on subgroups of the respiratory cycle has provided a pathway to classify various diseases or pathology that stem from airway obstruction.

Analysis of wheeze sounds during tidal breathing according to severity levels in asthma patients

Objective: This study aimed to statistically analyze the behavior of time-frequency features in digital recordings of wheeze sounds obtained from patients with various levels of asthma severity (mild, moderate, and severe), and this analysis was based on the auscultation location and/or breath phase. Method: Segmented and validated wheeze sounds were collected from the trachea and lower lung base (LLB) of 55 asthmatic patients during tidal breathing maneuvers and grouped into nine different datasets. The quartile frequencies F₂₅, F₅₀, F₇₅, F₉₀ and F₉₉, mean frequency (MF) and average power (AP) were computed as features, and a univariate statistical analysis was then performed to analyze the behavior of the time-frequency features. Results: All features generally showed statistical significance in most of the datasets for all severity levels [χ² = 6.021-71.65, p < 0.05, η² = 0.01-0.52]. Of the seven investigated features, only AP showed statistical significance in all the datasets. F₂₅, F₇₅, F₉₀ and F₉₉ exhibited statistical significance in at least six datasets [χ² = 4.852-65.63, p < 0.05, η² = 0.01-0.52], and F₂₅, F₅₀ and MF showed statistical significance with a large η² in all trachea-related datasets [χ² = 13.54-55.32, p < 0.05, η² = 0.13-0.33]. Conclusion: The results obtained for the time-frequency features revealed that (1) the asthma severity levels of patients can be identified through a set of selected features with tidal breathing, (2) tracheal wheeze sounds are more sensitive and specific predictors of severity levels and (3) inspiratory and expiratory wheeze sounds are almost equally informative.

Characterization and classification of asthmatic wheeze sounds according to severity level using spectral integrated features

OBJECTIVE

This study aimed to investigate and classify wheeze sounds of asthmatic patients according to their severity level (mild, moderate and severe) using spectral integrated (SI) features.

METHOD

Segmented and validated wheeze sounds were obtained from auscultation recordings of the trachea and lower lung base of 55 asthmatic patients during tidal breathing manoeuvres. The segments were multi-labelled into 9 groups based on the auscultation location and/or breath phases. Bandwidths were selected based on the physiology, and a corresponding SI feature was computed for each segment. Univariate and multivariate statistical analyses were then performed to investigate the discriminatory behaviour of the features with respect to the severity levels in the various groups. The asthmatic severity levels in the groups were then classified using the ensemble (ENS), support vector machine (SVM) and k-nearest neighbour (KNN) methods.

RESULTS AND CONCLUSION

All statistical comparisons exhibited a significant difference (p < 0.05) among the severity levels with few exceptions. In the classification experiments, the ensemble classifier exhibited better performance in terms of sensitivity, specificity and positive predictive value (PPV). The trachea inspiratory group showed the highest classification performance compared with all the other groups. Overall, the best PPV for the mild, moderate and severe samples were 95% (ENS), 88% (ENS) and 90% (SVM), respectively. With respect to location, the tracheal related wheeze sounds were most sensitive and specific predictors of asthma severity levels. In addition, the classification performances of the inspiratory and expiratory related groups were comparable, suggesting that the samples from these locations are equally informative.

Acoustic Methods for Pulmonary Diagnosis

Recent developments in sensor technology and computational analysis methods enable new strategies to measure and interpret lung acoustic signals that originate internally, such as breathing or vocal sounds, or are externally introduced, such as in chest percussion or airway insonification. A better understanding of these sounds has resulted in a new instrumentation that allows for highly accurate as well as portable options for measurement in the hospital, in the clinic, and even at home. This review outlines the instrumentation for acoustic stimulation and measurement of the lungs. We first review the fundamentals of acoustic lung signals and the pathophysiology of the diseases that these signals are used to detect. Then, we focus on different methods of measuring and creating signals that have been used in recent research for pulmonary disease diagnosis. These new methods, combined with signal processing and modeling techniques, lead to a reduction in noise and allow improved feature extraction and signal classification. We conclude by presenting the results of human subject studies taking advantage of both the instrumentation and signal processing tools to accurately diagnose common lung diseases. This paper emphasizes the active areas of research within modern lung acoustics and encourages the standardization of future work in this field.

Multichannel lung sound analysis for asthma detection

BACKGROUND AND OBJECTIVE

Lung sound signals convey valuable information of the lung status. Auscultation is an effective technique to appreciate the condition of the respiratory system using lung sound signals. The prior works on asthma detection from lung sound signals rely on the presence of wheeze. In this paper, we have classified normal and asthmatic subjects using advanced signal processing of posterior lung sound signals, even in the absence of wheeze.

METHODS

We collected lung sounds of 60 subjects (30 normal and 30 asthma) using a novel 4-channel data acquisition system from four different positions over the posterior chest, as suggested by the pulmonologist. A spectral subband based feature extraction scheme is proposed that works with artificial neural network (ANN) and support vector machine (SVM) classifiers for the multichannel signal. The power spectral density (PSD) is estimated from extracted lung sound cycle using Welch's method, which then decomposed into uniform subbands. A set of statistical features is computed from each subband and applied to ANN and SVM classifiers to classify normal and asthmatic subjects.

RESULTS

In the first part of this study, the performances of each individual channel and four channels together are evaluated where the combined channel performance is found superior to that of individual channels. Next, the performances of all possible combinations of the channels are investigated and the best classification accuracies of 89.2( ± 3.87)% and 93.3( ± 3.10)% are achieved for 2-channel and 3-channel combinations in ANN and SVM classifiers, respectively.

CONCLUSIONS

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. The channel combination study gives insight into the contribution of respective lung sound collection areas and their combinations in asthma detection.

Peripheral bronchial obstruction evaluation in patients with asthma by lung sound analysis and impulse oscillometry

BACKGROUND

Computer-aided lung sound analysis (LSA) has been reported to be useful for evaluating airway inflammation and obstruction in asthma patients. We investigated the relation between LSA and impulse oscillometry with the evaluation of peripheral airway obstruction.

METHODS

A total of 49 inhaled corticosteroid-naive bronchial asthma patients underwent LSA, spirometry, impulse oscillometry, and airway hyperresponsiveness testing. The data were analyzed to assess correlations between the expiration: inspiration lung sound power ratio (dB) at low frequencies between 100 and 195 Hz (E/I LF) and various parameters.

RESULTS

E/I LF and X5 were identified as independent factors that affect V˙_{50,%predicted}. E/I LF showed a positive correlation with R5 (r = 0.34, p = 0.017), R20 (r = 0.34, p = 0.018), reactance area (AX, r = 0.40, p = 0.005), and resonant frequency of reactance (Fres, r = 0.32, p = 0.024). A negative correlation was found between E/I LF and X5 (r = -0.47, p = 0.0006). E/I LF showed a negative correlation with FEV₁/FVC(%), FEV_1,%predicted, V˙_{50,%predicted}, and V˙_{25,%predicted} (r = -0.41, p = 0.003; r = -0.44, p = 0.002; r = -0.49, p = 0.0004; and r = -0.30, p = 0.024, respectively). E/I LF was negatively correlated with log PC₂₀ (r = -0.30, p = 0.024). Log PC₂₀, X5, and past smoking were identified as independent factors that affected E/I LF level.

CONCLUSIONS

E/I LF as with X5 can be an indicator of central and peripheral airway obstruction in bronchial asthma patients.

Artificial intelligence techniques used in respiratory sound analysis--a systematic review

Artificial intelligence (AI) has recently been established as an alternative method to many conventional methods. The implementation of AI techniques for respiratory sound analysis can assist medical professionals in the diagnosis of lung pathologies. This article highlights the importance of AI techniques in the implementation of computer-based respiratory sound analysis. Articles on computer-based respiratory sound analysis using AI techniques were identified by searches conducted on various electronic resources, such as the IEEE, Springer, Elsevier, PubMed, and ACM digital library databases. Brief descriptions of the types of respiratory sounds and their respective characteristics are provided. We then analyzed each of the previous studies to determine the specific respiratory sounds/pathology analyzed, the number of subjects, the signal processing method used, the AI techniques used, and the performance of the AI technique used in the analysis of respiratory sounds. A detailed description of each of these studies is provided. In conclusion, this article provides recommendations for further advancements in respiratory sound analysis.

Prediction of airway inflammation in patients with asymptomatic asthma by using lung sound analysis

BACKGROUND

The intensity and frequency of sounds in a lung sound analysis (LSA) may be related to airway constriction; however, whether any factors of an LSA can predict airway eosinophilic inflammation in patients with asthma is unknown.

OBJECTIVE

To determine whether an LSA can predict airway eosinophilic inflammation in patients with asymptomatic asthma.

METHODS

The expiratory-inspiratory ratios of sound power in the low-frequency range (E-I LF) from 36 patients with asymptomatic asthma were compared with those of 14 healthy controls. The relations of E-I LF with airway eosinophilic inflammation were analyzed. The E-I LF cutoff value for predicting airway eosinophilic inflammation also was analyzed.

RESULTS

The mean ± SD E-I LF was higher in the patients with asthma and with increased sputum eosinophils than in those patients without increased sputum eosinophils (0.45 ± 0.24 vs 0.20 ± 0.12; P < .001) or in the healthy controls (0.25 ± 0.10; P = .003). A multiple regression analysis showed that the sputum eosinophil ratio and exhaled nitric oxide were independently correlated with E-I LF, P = .0003 and P = .032, respectively. For the prediction of increased sputum eosinophils and increased fractional exhaled nitric oxide levels, the E-I LF thresholds of 0.29 and 0.30 showed sensitivities of 0.80 and 0.74 and specificities of 0.83 and 0.77, respectively.

CONCLUSIONS

We showed that LSAs can safely predict airway inflammation of patients with asymptomatic asthma.

The development of an automated device for asthma monitoring for adolescents: methodologic approach and user acceptability

BACKGROUND

Many adolescents suffer serious asthma related morbidity that can be prevented by adequate self-management of the disease. The accurate symptom monitoring by patients is the most fundamental antecedent to effective asthma management. Nonetheless, the adequacy and effectiveness of current methods of symptom self-monitoring have been challenged due to the individuals' fallible symptom perception, poor adherence, and inadequate technique. Recognition of these limitations led to the development of an innovative device that can facilitate continuous and accurate monitoring of asthma symptoms with minimal disruption of daily routines, thus increasing acceptability to adolescents.

OBJECTIVE

The objectives of this study were to: (1) describe the development of a novel symptom monitoring device for teenagers (teens), and (2) assess their perspectives on the usability and acceptability of the device.

METHODS

Adolescents (13-17 years old) with and without asthma participated in the evolution of an automated device for asthma monitoring (ADAM), which comprised three phases, including development (Phase 1, n=37), validation/user acceptability (Phase 2, n=84), and post hoc validation (Phase 3, n=10). In Phase 1, symptom algorithms were identified based on the acoustic analysis of raw symptom sounds and programmed into a popular mobile system, the iPod. Phase 2 involved a 7 day trial of ADAM in vivo, and the evaluation of user acceptance using an acceptance survey and individual interviews. ADAM was further modified and enhanced in Phase 3.

RESULTS

Through ADAM, incoming audio data were digitized and processed in two steps involving the extraction of a sequence of descriptive feature vectors, and the processing of these sequences by a hidden Markov model-based Viterbi decoder to differentiate symptom sounds from background noise. The number and times of detected symptoms were stored and displayed in the device. The sensitivity (true positive) of the updated cough algorithm was 70% (21/30), and, on average, 2 coughs per hour were identified as false positive. ADAM also kept track of the their activity level throughout the day using the mobile system's built in accelerometer function. Overall, the device was well received by participants who perceived it as attractive, convenient, and helpful. The participants recognized the potential benefits of the device in asthma care, and were eager to use it for their asthma management.

CONCLUSIONS

ADAM can potentially automate daily symptom monitoring with minimal intrusiveness and maximal objectivity. The users' acceptance of the device based on its recognized convenience, user-friendliness, and usefulness in increasing symptom awareness underscores ADAM's potential to overcome the issues of symptom monitoring including poor adherence, inadequate technique, and poor symptom perception in adolescents. Further refinement of the algorithm is warranted to improve the accuracy of the device. Future study is also needed to assess the efficacy of the device in promoting self-management and asthma outcomes.

Computerized lung sound analysis as diagnostic aid for the detection of abnormal lung sounds: a systematic review and meta-analysis

RATIONALE

The standardized use of a stethoscope for chest auscultation in clinical research is limited by its inherent inter-listener variability. Electronic auscultation and automated classification of recorded lung sounds may help prevent some of these shortcomings.

OBJECTIVE

We sought to perform a systematic review and meta-analysis of studies implementing computerized lung sound analysis (CLSA) to aid in the detection of abnormal lung sounds for specific respiratory disorders.

METHODS

We searched for articles on CLSA in MEDLINE, EMBASE, Cochrane Library and ISI Web of Knowledge through July 31, 2010. Following qualitative review, we conducted a meta-analysis to estimate the sensitivity and specificity of CLSA for the detection of abnormal lung sounds.

MEASUREMENTS AND MAIN RESULTS

Of 208 articles identified, we selected eight studies for review. Most studies employed either electret microphones or piezoelectric sensors for auscultation, and Fourier Transform and Neural Network algorithms for analysis and automated classification of lung sounds. Overall sensitivity for the detection of wheezes or crackles using CLSA was 80% (95% CI 72-86%) and specificity was 85% (95% CI 78-91%).

CONCLUSIONS

While quality data on CLSA are relatively limited, analysis of existing information suggests that CLSA can provide a relatively high specificity for detecting abnormal lung sounds such as crackles and wheezes. Further research and product development could promote the value of CLSA in research studies or its diagnostic utility in clinical settings.

Application of intelligent systems in asthma disease: designing a fuzzy rule-based system for evaluating level of asthma exacerbation

This paper discusses the capacities of artificial intelligence in the process of asthma diagnosing and asthma treatment. Developed intelligent systems for asthma disease have been classified in five categories including diagnosing, evaluating, management, communicative facilities, and prediction. Considering inputs, results, and methodologies of the systems show that by focusing on meticulous analysis of quality of life as an input variable and developing patient-based systems, under-diagnosing and asthma morbidity and mortality would decrease significantly. Regard to the importance of accurate evaluation in accurate prescription and expeditious treatment, the methodology of developing a fuzzy expert system for evaluating level of asthma exacerbation is presented in this paper too. The performance of this system has been tested in Asthma, Allergy, and Immunology Center of Iran using 25 asthmatic patients. Comparison between system's results and physicians' evaluations using Kappa coefficient (K) reinforces the value of K = 1. In addition this system assigns a degree in gradation (0-10) to every patient representing the slight differences between patients assigned to a specific category.

Classification of respiratory signals by linear analysis

The aim of this study is the classification of wheeze and non-wheeze epochs within respiratory sound signals acquired from patients with asthma and COPD. Since a wheeze signal, having a sinusoidal waveform, has a different behavior in time and frequency domains from that of a non-wheeze signal, the features selected for classification are kurtosis, Renyi entropy, f(50)/ f(90) ratio and mean-crossing irregularity. Upon calculation of these features for each wheeze and non-wheeze portion, the whole data scattered as two classes in four dimensional feature space is projected using Fisher Discriminant Analysis (FDA) onto the single dimensional space that separates the two classes best. Observing that the two classes are visually well separated in this new space, Neyman-Pearson hypothesis testing is applied. Finally, the correct classification rate is %95.1 for the training set, and leave-one-out approach pursuing the above methodology yields a success rate of %93.5 for the test set.

Pattern recognition methods applied to respiratory sounds classification into normal and wheeze classes

In this paper, we present the pattern recognition methods proposed to classify respiratory sounds into normal and wheeze classes. We evaluate and compare the feature extraction techniques based on Fourier transform, linear predictive coding, wavelet transform and Mel-frequency cepstral coefficients (MFCC) in combination with the classification methods based on vector quantization, Gaussian mixture models (GMM) and artificial neural networks, using receiver operating characteristic curves. We propose the use of an optimized threshold to discriminate the wheezing class from the normal one. Also, post-processing filter is employed to considerably improve the classification accuracy. Experimental results show that our approach based on MFCC coefficients combined to GMM is well adapted to classify respiratory sounds in normal and wheeze classes. McNemar's test demonstrated significant difference between results obtained by the presented classifiers (p<0.05).

Respiratory sounds classification using cepstral analysis and Gaussian mixture models

The Cepstral analysis is proposed with Gaussian Mixture Models (GMM) method to classify respiratory sounds in two categories: normal and wheezing. The sound signal is divided in overlapped segments, which are characterized by a reduced dimension feature vectors using Mel-Frequency Cepstral Coefficients (MFCC) or subband based Cepstral parameters (SBC). The proposed schema is compared with other classifiers: Vector Quantization (VQ) and Multi-Layer Perceptron (MLP) neural networks. A post processing is proposed to improve the classification results.

Biomedical informatics applications for asthma care: a systematic review

Asthma is a common condition associated with significant patient morbidity and health care costs. Although widely accepted evidence-based guidelines for asthma management exist, unnecessary variation in patient care remains. Application of biomedical informatics techniques is one potential way to improve care for asthmatic patients. We performed a systematic literature review to identify computerized applications for clinical asthma care. Studies were evaluated for their clinical domain, developmental stage and study design. Additionally, prospective trials were identified and analyzed for potential study biases, study effects, and clinical study characteristics. Sixty-four papers were selected for review. Publications described asthma detection or diagnosis (18 papers), asthma monitoring or prevention (13 papers), patient education (13 papers), and asthma guidelines or therapy (20 papers). The majority of publications described projects in early stages of development or with non-prospective study designs. Twenty-one prospective trials were identified, which evaluated both clinical and non-clinical impacts on patient care. Most studies took place in the outpatient clinic environment, with minimal study of the emergency department or inpatient settings. Few studies demonstrated evidence of computerized applications improving clinical outcomes. Further research is needed to prospectively evaluate the impact of using biomedical informatics to improve care of asthmatic patients.

Combining neural network and genetic algorithm for prediction of lung sounds

Recognition of lung sounds is an important goal in pulmonary medicine. In this work, we present a study for neural networks-genetic algorithm approach intended to aid in lung sound classification. Lung sound was captured from the chest wall of The subjects with different pulmonary diseases and also from the healthy subjects. Sound intervals with duration of 15-20 s were sampled from subjects. From each interval, full breath cycles were selected. Of each selected breath cycle, a 256-point Fourier Power Spectrum Density (PSD) was calculated. Total of 129 data values calculated by the spectral analysis are selected by genetic algorithm and applied to neural network. Multilayer perceptron (MLP) neural network employing backpropagation training algorithm was used to predict the presence or absence of adventitious sounds (wheeze and crackle). We used genetic algorithms to search for optimal structure and training parameters of neural network for a better predicting of lung sounds. This application resulted in designing of optimum network structure and, hence reducing the processing load and time.

Neural classification of lung sounds using wavelet coefficients

Electronic auscultation is an efficient technique to evaluate the condition of respiratory system using lung sounds. As lung sound signals are non-stationary, the conventional method of frequency analysis is not highly successful in diagnostic classification. This paper deals with a novel method of analysis of lung sound signals using wavelet transform, and classification using artificial neural network (ANN). Lung sound signals were decomposed into the frequency subbands using wavelet transform and a set of statistical features was extracted from the subbands to represent the distribution of wavelet coefficients. An ANN based system, trained using the resilient back propagation algorithm, was implemented to classify the lung sounds to one of the six categories: normal, wheeze, crackle, squawk, stridor, or rhonchus.

Comparison of neural network predictors in the classification of tracheal-bronchial breath sounds by respiratory auscultation

Despite extensive research in the area of identification and discrimination of tracheal-bronchial breath sounds by computer analysis, the process of identifying auscultated sounds is still subject to high estimation uncertainties. Here we assess the performance of the relatively new constructive probabilistic neural network (CPNN) against the more common classifiers, namely the multilayer perceptron (MLP) and radial basis function network (RBFN), in classifying a broad range of tracheal-bronchial breath sounds. We present our data as signal estimation models of the tracheal-bronchial frequency spectra. We have examined the trained structure of the CPNN with respect to the other architectures and conclude that this architecture offers an attractive means with which to analyse this type of data. This is based partly on the classification accuracies attained by the CPNN, MLP and RBFN which were 97.8, 77.8 and 96.2%, respectively. We concluded that CPNN and RBFN networks are capable of working successfully with this data, with these architectures being acceptable in terms of topological size and computational overhead requirements. We further believe that the CPNN is an attractive classification mechanism for auscultated data analysis due to its optimal data model generation properties and computationally lightweight architecture.