Lung Sound Analysis Research Articles

MEMS and ECM Sensor Technologies for Cardiorespiratory Sound Monitoring-A Comprehensive Review.

This paper presents a comprehensive review of cardiorespiratory auscultation sensing devices (i.e., stethoscopes), which is useful for understanding the theoretical aspects and practical design notes. In this paper, we first introduce the acoustic properties of the heart and lungs, as well as a brief history of stethoscope evolution. Then, we discuss the basic concept of electret condenser microphones (ECMs) and a stethoscope based on them. Then, we discuss the microelectromechanical systems (MEMSs) technology, particularly focusing on piezoelectric transducer sensors. This paper comprehensively reviews sensing technologies for cardiorespiratory auscultation, emphasizing MEMS-based wearable designs in the past decade. To our knowledge, this is the first paper to summarize ECM and MEMS applications for heart and lung sound analysis.

Empowering Healthcare: TinyML for Precise Lung Disease Classification

Respiratory diseases such as asthma pose significant global health challenges, necessitating efficient and accessible diagnostic methods. The traditional stethoscope is widely used as a non-invasive and patient-friendly tool for diagnosing respiratory conditions through lung auscultation. However, it has limitations, such as a lack of recording functionality, dependence on the expertise and judgment of physicians, and the absence of noise-filtering capabilities. To overcome these limitations, digital stethoscopes have been developed to digitize and record lung sounds. Recently, there has been growing interest in the automated analysis of lung sounds using Deep Learning (DL). Nevertheless, the execution of large DL models in the cloud often leads to latency, dependency on internet connectivity, and potential privacy issues due to the transmission of sensitive health data. To address these challenges, we developed Tiny Machine Learning (TinyML) models for the real-time detection of respiratory conditions by using lung sound recordings, deployable on low-power, cost-effective devices like digital stethoscopes. We trained three machine learning models—a custom CNN, an Edge Impulse CNN, and a custom LSTM—on a publicly available lung sound dataset. Our data preprocessing included bandpass filtering and feature extraction through Mel-Frequency Cepstral Coefficients (MFCCs). We applied quantization techniques to ensure model efficiency. The custom CNN model achieved the highest performance, with 96% accuracy and 97% precision, recall, and F1-scores, while maintaining moderate resource usage. These findings highlight the potential of TinyML to provide accessible, reliable, and real-time diagnostic tools, particularly in remote and underserved areas, demonstrating the transformative impact of integrating advanced AI algorithms into portable medical devices. This advancement facilitates the prospect of automated respiratory health screening using lung sounds.

LungNeXt: A novel lightweight network utilizing enhanced mel-spectrogram for lung sound classification

Lung auscultation is essential for early lung condition detection. Categorizing adventitious lung sounds requires expert discrimination by medical specialists. This paper details the features of LungNeXt, a novel classification model specifically designed for lung sound analysis. Furthermore, we propose two auxiliary methods: RandClipMix (RCM) for data augmentation and Enhanced Mel-Spectrogram for Feature Extraction (EMFE). RCM addresses the issue of data imbalance by randomly mixing clips within the same category to create new adventitious lung sounds. EMFE augments specific frequency bands in spectrograms to highlight adventitious features. These contributions enable LungNeXt to achieve outstanding performance. LungNeXt optimally integrates an appropriate number of NeXtblocks, ensuring superior performance and a lightweight model architecture. The proposed RCM and EMFE methods, along with the LungNeXt classification network, have been evaluated on the SPRSound dataset. Experimental results revealed a commendable score of 0.5699 for the lung sound five-category task on SPRSound. Specifically, the LungNeXt model is characterized by its efficiency, with only 3.804M parameters and a computational complexity of 0.659G FLOPS. This lightweight and efficient model is particularly well-suited for applications in electronic stethoscope back-end processing equipment, providing efficient diagnostic advice to physicians and patients.

Automated Interpretation of Lung Sounds by Deep Learning in Children With Asthma: Scoping Review and Strengths, Weaknesses, Opportunities, and Threats Analysis.

The interpretation of lung sounds plays a crucial role in the appropriate diagnosis and management of pediatric asthma. Applying artificial intelligence (AI) to this task has the potential to better standardize assessment and may even improve its predictive potential. This study aims to objectively review the literature on AI-assisted lung auscultation for pediatric asthma and provide a balanced assessment of its strengths, weaknesses, opportunities, and threats. A scoping review on AI-assisted lung sound analysis in children with asthma was conducted across 4 major scientific databases (PubMed, MEDLINE Ovid, Embase, and Web of Science), supplemented by a gray literature search on Google Scholar, to identify relevant studies published from January 1, 2000, until May 23, 2023. The search strategy incorporated a combination of keywords related to AI, pulmonary auscultation, children, and asthma. The quality of eligible studies was assessed using the ChAMAI (Checklist for the Assessment of Medical Artificial Intelligence). The search identified 7 relevant studies out of 82 (9%) to be included through an academic literature search, while 11 of 250 (4.4%) studies from the gray literature search were considered but not included in the subsequent review and quality assessment. All had poor to medium ChAMAI scores, mostly due to the absence of external validation. Identified strengths were improved predictive accuracy of AI to allow for prompt and early diagnosis, personalized management strategies, and remote monitoring capabilities. Weaknesses were the heterogeneity between studies and the lack of standardization in data collection and interpretation. Opportunities were the potential of coordinated surveillance, growing data sets, and new ways of collaboratively learning from distributed data. Threats were both generic for the field of medical AI (loss of interpretability) but also specific to the use case, as clinicians might lose the skill of auscultation. To achieve the opportunities of automated lung auscultation, there is a need to address weaknesses and threats with large-scale coordinated data collection in globally representative populations and leveraging new approaches to collaborative learning.

The analysis of lung sounds in infants and children with a history of wheezing/asthma using an automatic procedure

BackgroundLung sound analysis parameters have been reported to be useful biomarkers for evaluating airway condition. We developed an automatic lung sound analysis software program for infants and children based on lung sound spectral curves of frequency and power by leveraging machine learning (ML) technology.MethodsTo put this software program into clinical practice, in Study 1, the reliability and reproducibility of the software program using data from younger children were examined. In Study 2, the relationship between lung sound parameters and respiratory flow (L/s) was evaluated using data from older children. In Study 3, we conducted a survey using the ATS-DLD questionnaire to evaluate the clinical usefulness. The survey focused on the history of wheezing and allergies, among healthy 3-year-old infants, and then measured lung sounds. The clinical usefulness was evaluated by comparing the questionnaire results with the results of the new lung sound parameters.ResultsIn Studies 1 and 2, the parameters of the new software program demonstrated excellent reproducibility and reliability, and were not affected by airflow (L/s). In Study 3, infants with a history of wheezing showed lower FAP0 and RPF75p (p < 0.001 and p = 0.025, respectively) and higher PAP0 (p = 0.001) than healthy infants. Furthermore, infants with asthma/asthma-like bronchitis showed lower FAP0 (p = 0.002) and higher PAP0 (p = 0.001) than healthy infants.ConclusionsLung sound parameters obtained using the ML algorithm were able to accurately assess the respiratory condition of infants. These parameters are useful for the early detection and intervention of childhood asthma.

Exploring the link between a novel approach for computer aided lung sound analysis and imaging biomarkers: a cross-sectional study

BackgroundComputer Aided Lung Sound Analysis (CALSA) aims to overcome limitations associated with standard lung auscultation by removing the subjective component and allowing quantification of sound characteristics. In this proof-of-concept study, a novel automated approach was evaluated in real patient data by comparing lung sound characteristics to structural and functional imaging biomarkers.MethodsPatients with cystic fibrosis (CF) aged > 5y were recruited in a prospective cross-sectional study. CT scans were analyzed by the CF-CT scoring method and Functional Respiratory Imaging (FRI). A digital stethoscope was used to record lung sounds at six chest locations. Following sound characteristics were determined: expiration-to-inspiration (E/I) signal power ratios within different frequency ranges, number of crackles per respiratory phase and wheeze parameters. Linear mixed-effects models were computed to relate CALSA parameters to imaging biomarkers on a lobar level.Results222 recordings from 25 CF patients were included. Significant associations were found between E/I ratios and structural abnormalities, of which the ratio between 200 and 400 Hz appeared to be most clinically relevant due to its relation with bronchiectasis, mucus plugging, bronchial wall thickening and air trapping on CT. The number of crackles was also associated with multiple structural abnormalities as well as regional airway resistance determined by FRI. Wheeze parameters were not considered in the statistical analysis, since wheezing was detected in only one recording.ConclusionsThe present study is the first to investigate associations between auscultatory findings and imaging biomarkers, which are considered the gold standard to evaluate the respiratory system. Despite the exploratory nature of this study, the results showed various meaningful associations that highlight the potential value of automated CALSA as a novel non-invasive outcome measure in future research and clinical practice.

Approaches for Respiratory Sound Analysis in Identification of Respiratory Diseases

Purpose: Medical professionals throughout the world prefer to use conventional stethoscopes to listen to respiratory sounds. Listening to respiratory sounds through stethoscopes is a subjective matter, and proper diagnosis of the disease depends on the skills and ability of the doctor. Computerized analysis of respiratory sounds can help doctors and researchers to characterize different abnormal respiratory patterns and make informed decisions.&#x0D; Materials and Methods: This study includes previously reported work in different normal and abnormal respiratory sounds. The IEEE, PubMed, Google Scholar and Elsevier databases were searched and studies with the keywords of lung sound analysis, respiratory sound analysis, and respiratory sound classification were included. Detailed characteristics of normal and abnormal respiratory sounds are mentioned. In addition, Time-amplitude characteristics of different respiratory sound plots are obtained using MATLAB and ICBHI database. This study systematically discusses different approaches for respiratory sound analysis like visual analysis of the time-amplitude signals, frequency analysis, and spectral analysis using fast Fourier transform, statistical analysis, and machine learning approach. A list of relevant datasets is mentioned that can help researchers to do further analysis in this domain.&#x0D; Results: The careful observations and analysis show the possibility of predicting respiratory diseases by extracting suitable parameters such as the frequency response and spectral characteristics of the signal. Power spectral density can help us to calculate the maximum, median frequency over an extended period. Using machine learning we can estimate the energy, entropy, spectral features, and wavelets of the signals.&#x0D; Conclusion: Computer-based respiratory sound analysis can help medical professionals in making informed decisions. This will help in early diagnosis and devise effective treatment plans for the patients.

Lung sound analysis for predicting recurrent wheezing in preschool children

BackgroundIn young healthy children, assessing airflow limitation may be difficult because of the narrowing of the airways, which is a pathology of asthma, and responsiveness to bronchodilators. ObjectiveWe investigated whether lung sound analysis could predict the development of recurrent wheezing (RW), one of the signs of asthma. MethodsIn healthy children aged 3 to 24 months, we recorded and analyzed lung sounds before and after inhalation of bronchodilators and conducted a questionnaire survey. The children were followed up and assessed for the development of RW at 3 years of age. ResultsOf the 62 patients analyzed, 19 (30.6%) developed RW. The parameters RPF50 and RPF75, calculated by lung sound analysis, were lower in the RW group, with odds ratios (95% CI) of 0.77 (0.61–0.98) and 0.81 (0.66–0.99), respectively. The change rate of lung sound analysis parameters after bronchodilator inhalation was not different among all participants; however, in the subgroup of children with a history of atopic dermatitis, B4/AT and ΔRPF50 were elevated in the RW group (P = .015 and P = .041, respectively). In the subgroup of children with total IgE >20 kUA/L, the sensitivity and specificity (95% CI) for predicting the development of RW were 85.7% (48.7-99.3) and 68.8% (44.4-85.8), respectively, when the cutoff value of ΔRPF50 was set at 10.5%. ConclusionThe lung sound analysis method allows noninvasive assessment of the airway, including airway hypersensitivity, in young children and may predict the risk of developing RW.

Multiple Channels Model Based on Mel Spectrogram for Classifying Abnormalities in Lung Sound

Lung sound analysis plays an important role in the assessment and diagnosis of respiratory conditions and diseases. It can provide valuable information about the functioning of the respiratory system, including the airways, lungs, and associated structures. By analyzing the characteristics of lung sounds, healthcare professionals can gain insights into the presence of abnormalities, such as airway obstructions, lung diseases, and respiratory infections. In this paper, a multiple channel model for processing and classifying abnormalities in lung sound is proposed, which utilize the characteristics of Mel spectrogram and the Empirical Mode Decomposition (EMD). Unlike previous research which directly convert the lung sound into scalogram or spectrogram, the pre-processing of the original audio signal is considered and focused in this paper. This pre-processing step includes denoising, resampling, padding and augmentation, which incredibly increase the quality of the input signal. Finally, the multiple channel is put into the VGG16 deep learning model to classify the abnormalities in lung sound, including wheezes, crackles, and both. The model is trained and tested on the benchmark ICBHI dataset. The proposed model has shown better performance when compared with the state-of-the-art researches.

A dual-purpose deep learning model for auscultated lung and tracheal sound analysis based on mixed set training

Many deep learning–based computerized respiratory sound analysis methods have previously been developed. However, these studies focus on either lung sound only or tracheal sound only. The effectiveness of using a lung sound analysis algorithm on tracheal sound and vice versa is rarely reported. Furthermore, no one knows whether using lung and tracheal sounds together in training a deep learning-based respiratory sound analysis model is beneficial. In this study, we first constructed a tracheal sound database, HF_Tracheal_V1, containing 10,448 15-s tracheal sound recordings, 21,741 inhalation labels, 15,858 exhalation labels, and 6414 continuous adventitious sound (CAS) labels collected from 227 participants undergoing a diagnostic/surgical procedure under monitored anesthesia care. HF_Tracheal_V1 and our previously built lung sound database, HF_Lung_V2, were either combined (mixed set), used one after the other (domain adaptation), or used alone to train convolutional neural network bidirectional gate recurrent unit models for inhalation, exhalation, and CAS detection in lung and tracheal sounds. The results revealed that the models trained using lung sound alone performed poorly in tracheal sound analysis and vice versa. However, mixed set training or domain adaptation improved the performance for 1) inhalation and exhalation detection in lung sounds and 2) inhalation, exhalation, and CAS detection in tracheal sounds compared to positive controls (the models trained using lung sound alone and used in lung sound analysis and vice versa). In particular, the model trained on the mixed set had great flexibility to serve two purposes, lung and tracheal sound analyses, at the same time.

A Multi-Center Clinical Trial for Wireless Stethoscope-Based Diagnosis and Prognosis of Children Community-Acquired Pneumonia.

Community-Acquired Pneumonia (CAP) is a significant cause of child mortality globally, due to the lack of ubiquitous monitoring methods. Clinically, the wireless stethoscope can be a promising solution since lung sounds with crackles and tachypnea are considered as the typical symptoms of CAP. In this paper, we carried out a multi-center clinical trial in four hospitals to investigate the feasibility of using a wireless stethoscope for children CAP diagnosis and prognosis. The trial collects both the left and right lung sounds from children with CAP at the time of diagnosis, improvement, and recovery. A bilateral pulmonary audio-auxiliary model (BPAM) is proposed for lung sound analysis. It learns the underlying pathological paradigm for the CAP classification by mining the contextual information of audio while preserving the structured information of breathing cycle. The clinical validation shows that the specificity and sensitivity of BPAM are over 92% in both the CAP diagnosis and prognosis for the subject-dependent experiment, over 50% in CAP diagnosis and 39% in CAP prognosis for the subject-independent experiment. Almost all benchmarked methods have improved performance by fusing left and right lung sounds, indicating the direction of hardware design and algorithmic improvement.

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.

Acoustic-Based Deep Learning Architectures for Lung Disease Diagnosis: A Comprehensive Overview.

Lung auscultation has long been used as a valuable medical tool to assess respiratory health and has gotten a lot of attention in recent years, notably following the coronavirus epidemic. Lung auscultation is used to assess a patient's respiratory role. Modern technological progress has guided the growth of computer-based respiratory speech investigation, a valuable tool for detecting lung abnormalities and diseases. Several recent studies have reviewed this important area, but none are specific to lung sound-based analysis with deep-learning architectures from one side and the provided information was not sufficient for a good understanding of these techniques. This paper gives a complete review of prior deep-learning-based architecture lung sound analysis. Deep-learning-based respiratory sound analysis articles are found in different databases including the Plos, ACM Digital Libraries, Elsevier, PubMed, MDPI, Springer, and IEEE. More than 160 publications were extracted and submitted for assessment. This paper discusses different trends in pathology/lung sound, the common features for classifying lung sounds, several considered datasets, classification methods, signal processing techniques, and some statistical information based on previous study findings. Finally, the assessment concludes with a discussion of potential future improvements and recommendations.

Effect of wheeze and lung function on lung sound parameters in children with asthma

BackgroundIn children with asthma, there are many cases in which wheeze is confirmed by auscultation with a normal lung function, or in which the lung function is decreased without wheeze. Using an objective lung sound analysis, we examined the effect of wheeze and the lung function on lung sound parameters in children with asthma. MethodsA total of 114 children with asthma (males to females = 80: 34, median age 10 years old) were analyzed for their lung sound parameters using conventional methods, and wheeze and the lung function were checked. The effects of wheeze and the lung function on lung sound parameters were examined. ResultsThe patients with wheeze or decreased forced expiratory flow and volume in 1 s (FEV1) (% pred) showed a significantly higher sound power of respiration and expiration-to-inspiration sound power ratio (E/I) than those without wheeze and a normal FEV1 (% pred). There was no marked difference in the sound power of respiration or E/I between the patients without wheeze and a decreased FEV1 (% pred) and the patients with wheeze and a normal FEV1 (% pred). ConclusionsOur data suggest that bronchial constriction in the asthmatic children with wheeze similarly exists in the asthmatic children with a decreased lung function. A lung sound analysis is likely to enable an accurate understanding of airway conditions.

A machine learning approach to the development and prospective evaluation of a pediatric lung sound classification model

Auscultation, a cost-effective and non-invasive part of physical examination, is essential to diagnose pediatric respiratory disorders. Electronic stethoscopes allow transmission, storage, and analysis of lung sounds. We aimed to develop a machine learning model to classify pediatric respiratory sounds. Lung sounds were digitally recorded during routine physical examinations at a pediatric pulmonology outpatient clinic from July to November 2019 and labeled as normal, crackles, or wheezing. Ensemble support vector machine models were trained and evaluated for four classification tasks (normal vs. abnormal, crackles vs. wheezing, normal vs. crackles, and normal vs. wheezing) using K-fold cross-validation (K = 10). Model performance on a prospective validation set (June to July 2021) was compared with those of pediatricians and non-pediatricians. Total 680 clips were used for training and internal validation. The model accuracies during internal validation for normal vs. abnormal, crackles vs. wheezing, normal vs. crackles, and normal vs. wheezing were 83.68%, 83.67%, 80.94%, and 90.42%, respectively. The prospective validation (n = 90) accuracies were 82.22%, 67.74%, 67.80%, and 81.36%, respectively, which were comparable to pediatrician and non-pediatrician performance. An automated classification model of pediatric lung sounds is feasible and maybe utilized as a screening tool for respiratory disorders in this pandemic era.

Clinical application of a lung sound analysis in infants with respiratory syncytial virus acute bronchiolitis.

Objective investigation of the characteristics of acute bronchiolitis in infants is important for its diagnosis and treatment. Lung sound data of 50 patients diagnosed with respiratory syncytial virus (RSV) acute bronchiolitis (m:f = 29:21, median of age 7 months), 20 patients with RSV acute respiratory tract infections without acute bronchiolitis (m:f = 10:10, 5 months) and 38 age-matched control infants (m:f = 23:15, 8 months) were analyzed using a conventional method and compared. Furthermore, the relationships between lung sound parameters and clinical symptoms (clinical score, length of hospital stay and SpO2 level) in the bronchiolitis and the non-bronchiolitis patients were examined. Results of lung sound analysis showed that the inspiratory sound power of patients with RSV respiratory tract infections was low and the expiratory sound power was high compared with those of the controls. When the patients with RSV respiratory tract infections were divided into the bronchiolitis and non-bronchiolitis groups, the expiratory/inspiratory ratio of the bronchiolitis patients was greater than that of the non-bronchiolitis patients. There was no difference in the clinical symptoms, clinical score and length of hospital stay between the bronchiolitis and non-bronchiolitis patients, except for the SpO2 level on admission. Lung sound analysis confirmed that patients with RSV acute bronchiolitis present with marked airway narrowing. Considering these results as a characteristic of acute bronchiolitis, it would be meaningful to reflect it in the improvement of diagnosis, treatment and subsequent management.

Continuous Monitoring Versus Intermittent Auscultation of Wheezes in Patients Presenting With Acute Respiratory Distress

BackgroundAuscultation for an extended period of time using a wearable stethoscope enables objective computerized analysis and longitudinal assessment of lung sounds. However, this auscultation method differs from bedside auscultation in that clinicians are not present to optimize the quality of auscultation. No prior studies have compared these two auscultation methods. ObjectiveThe aim of this study was to compare intermittent auscultation using a conventional stethoscope with continuous auscultation using a wearable stethoscope for wheeze detection in patients who present with acute respiratory distress. MethodsPatients presenting to the emergency department with acute respiratory distress were enrolled. The Strados Remote Electronic Stethoscope Platform (RESP™) was used for continuous auscultation, and intermittent auscultation was performed using a U.S. Food and Drug Administration–cleared electronic stethoscope. A recording was made with an electronic stethoscope. Subsequently, continuous recording was made using RESP™, which continued until the patient was admitted or discharged from the emergency department. The number of captured wheezes in each recording was counted and validated by two board-certified physicians. ResultsFrom May 2018 to May 2019, 43 patients were enrolled in the study. Three patients were excluded from analysis due to incomplete audio recording data. The mean length of recording was 62.3 min for continuous auscultation and 0.7 min for intermittent auscultation; 77.5% (31 of 40) of intermittent recordings contained wheezes, in contrast to 85% (34 of 40) of continuous recordings. ConclusionsExtending the duration of auscultation using a wearable stethoscope in a noisy clinical environment showed comparable performance to standard of care intermittent auscultation in identifying patients who have wheezes.

Lung Sound Classification Using Co-Tuning and Stochastic Normalization.

Computational methods for lung sound analysis are beneficial for computer-aided diagnosis support, storage and monitoring in critical care. In this paper, we use pre-trained ResNet models as backbone architectures for classification of adventitious lung sounds and respiratory diseases. The learned representation of the pre-trained model is transferred by using vanilla fine-tuning, co-tuning, stochastic normalization and the combination of the co-tuning and stochastic normalization techniques. Furthermore, data augmentation in both time domain and time-frequency domain is used to account for the class imbalance of the ICBHI and our multi-channel lung sound dataset. Additionally, we introduce spectrum correction to account for the variations of the recording device properties on the ICBHI dataset. Empirically, our proposed systems mostly outperform all state-of-the-art lung sound classification systems for the adventitious lung sounds and respiratory diseases of both datasets.

A Progressively Expanded Database for Automated Lung Sound Analysis: An Update

We previously established an open-access lung sound database, HF_Lung_V1, and developed deep learning models for inhalation, exhalation, continuous adventitious sound (CAS), and discontinuous adventitious sound (DAS) detection. The amount of data used for training contributes to model accuracy. In this study, we collected larger quantities of data to further improve model performance and explored issues of noisy labels and overlapping sounds. HF_Lung_V1 was expanded to HF_Lung_V2 with a 1.43× increase in the number of audio files. Convolutional neural network–bidirectional gated recurrent unit network models were trained separately using the HF_Lung_V1 (V1_Train) and HF_Lung_V2 (V2_Train) training sets. These were tested using the HF_Lung_V1 (V1_Test) and HF_Lung_V2 (V2_Test) test sets, respectively. Segment and event detection performance was evaluated. Label quality was assessed. Overlap ratios were computed between inhalation, exhalation, CAS, and DAS labels. The model trained using V2_Train exhibited improved performance in inhalation, exhalation, CAS, and DAS detection on both V1_Test and V2_Test. Poor CAS detection was attributed to the quality of CAS labels. DAS detection was strongly influenced by the overlapping of DAS with inhalation and exhalation. In conclusion, collecting greater quantities of lung sound data is vital for developing more accurate lung sound analysis models.

Digital auscultation as a diagnostic aid to detect childhood pneumonia: A systematic review

BackgroundFrontline health care workers use World Health Organization Integrated Management of Childhood Illnesses (IMCI) guidelines for child pneumonia care in low-resource settings. IMCI guideline pneumonia diagnostic criterion performs with low specificity, resulting in antibiotic overtreatment. Digital auscultation with automated lung sound analysis may improve the diagnostic performance of IMCI pneumonia guidelines. This systematic review aims to summarize the evidence on detecting adventitious lung sounds by digital auscultation with automated analysis compared to reference physician acoustic analysis for child pneumonia diagnosis.MethodsIn this review, articles were searched from MEDLINE, Embase, CINAHL Plus, Web of Science, Global Health, IEEExplore database, Scopus, and the ClinicalTrial.gov databases from the inception of each database to October 27, 2021, and reference lists of selected studies and relevant review articles were searched manually. Studies reporting diagnostic performance of digital auscultation and/or computerized lung sound analysis compared against physicians’ acoustic analysis for pneumonia diagnosis in children under the age of 5 were eligible for this systematic review. Retrieved citations were screened and eligible studies were included for extraction. Risk of bias was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) tool. All these steps were independently performed by two authors and disagreements between the reviewers were resolved through discussion with an arbiter. Narrative data synthesis was performed.ResultsA total of 3801 citations were screened and 46 full-text articles were assessed. 10 studies met the inclusion criteria. Half of the studies used a publicly available respiratory sound database to evaluate their proposed work. Reported methodologies/approaches and performance metrics for classifying adventitious lung sounds varied widely across the included studies. All included studies except one reported overall diagnostic performance of the digital auscultation/computerised sound analysis to distinguish adventitious lung sounds, irrespective of the disease condition or age of the participants. The reported accuracies for classifying adventitious lung sounds in the included studies varied from 66.3% to 100%. However, it remained unclear to what extent these results would be applicable for classifying adventitious lung sounds in children with pneumonia.ConclusionsThis systematic review found very limited evidence on the diagnostic performance of digital auscultation to diagnose pneumonia in children. Well-designed studies and robust reporting are required to evaluate the accuracy of digital auscultation in the paediatric population.