Respiratory sound classification utilizing human auditory-based feature extraction

The Registry of the International Society for Heart and Lung Transplantation: Thirty-third Adult Lung and Heart–Lung Transplant Report—2016; Focus Theme: Primary Diagnostic Indications for Transplant

Pulmonary Hypertension: From an Orphan Disease to a Public Health Problem

Compared to human speech and multimedia audio, the amplitude of lung respiratory sounds is extremely weak, with minimal differences between the respiratory sounds of various lung diseases. Traditional methods, such as Mel‐frequency cepstral coefficients (MFCCs) and the Fourier transform, struggle to accurately extract respiratory sound characteristics of different lung diseases. Therefore, this article innovatively employs an improved C 0 complexity method to analyze the subtle differences in respiratory sounds, identifying the optimal tuning factor value that captures more local information and better characterizes the nonlinear stochastic properties of respiratory sounds. Subsequently, the second component of the MFCC was extracted and fused with the improved C 0 complexity and short‐term energy to propose a novel Mel‐energy complexity (MEC) feature. Finally, the fuzzy C‐means (FCM) clustering method was employed to process the MEC features, detecting the periodic endpoints of respiratory sounds, specifically, the endpoints of respiratory sounds during inhalation and exhalation, thereby laying a technical foundation for further research on respiratory sound recognition and visual management. To evaluate performance, we established a clinically collected respiratory sound database containing respiratory sounds from 154 individuals across four categories, including asthma and chronic obstructive pulmonary disease (COPD), with sounds recorded from eight chest locations per participant. The proposed method achieved average accuracy and F1 scores of 71% and 72%, respectively, outperforming multiple respiratory sound endpoint detection (RSED) approaches while demonstrating high robustness, particularly in processing pulmonary crackles.

Respiratory diseases are major health concerns worldwide. Chronic respiratory diseases (CRDs) are the third leading cause of death worldwide and some of the most common are chronic obstructive pulmonary disease (COPD), asthma, occupational lung diseases, and pulmonary hypertension. Despite having different etiology and characteristics, these diseases share several features, such as a persistent inflammatory state, chronic oxidative stress, impaired mucociliary clearance, and increased alveolar surface tension. CRDs are not curable; however, various forms of treatment, that help restore airway patency and reduce shortness of breath, can improve daily life for people living with these conditions. In this regard myo-inositol may represent a valid therapeutic adjuvant approach due to its properties. Being a redox balancer, an inflammation modulator, and, most importantly, a component of pulmonary surfactant, it may improve lung function and counteract symptoms associated with respiratory diseases, as recently evidenced in patients with COPD, COVID-19, asthma, and bronchiectasis. The aim of this review is to evaluate the potential therapeutic role of myo-inositol supplementation in the management of patients with respiratory diseases.

This work addresses the problem of lung sound classification, in particular, the problem of distinguishing between wheeze and normal sounds. Wheezing sound detection is an important step to associate lung sounds with an abnormal state of the respiratory system, usually associated with tuberculosis or another chronic obstructive pulmonary diseases (COPD). The paper presents an approach for automatic lung sound classification, which uses different state-of-the-art sound features in combination with a C-weighted support vector machine (SVM) classifier that works better for unbalanced data. Feature extraction methods used here are commonly applied in speech recognition and related problems thanks to the fact that they capture the most informative spectral content from the original signals. The evaluated methods were: Fourier transform (FT), wavelet decomposition using Wavelet Packet Transform bank of filters (WPT) and Mel Frequency Cepstral Coefficients (MFCC). For comparison, we evaluated and contrasted the proposed approach against previous works using different combination of features and/or classifiers. The different methods were evaluated on a set of lung sounds including normal and wheezing sounds. A leave-two-out per-case cross-validation approach was used, which, in each fold, chooses as validation set a couple of cases, one including normal sounds and the other including wheezing sounds. Experimental results were reported in terms of traditional classification performance measures: sensitivity, specificity and balanced accuracy. Our best results using the suggested approach, C-weighted SVM and MFCC, achieve a 82.1% of balanced accuracy obtaining the best result for this problem until now. These results suggest that supervised classifiers based on kernel methods are able to learn better models for this challenging classification problem even using the same feature extraction methods.

Chronic Obstructive Pulmonary Disease (COPD) is the top cause of death worldwide and a growing health care burden. Advanced COPD often causes pulmonary hypertension, affecting around 20% of patients. Severe PAH in COPD patients reduces median survival by 40 months. Early pulmonary hypertension identification in COPD patients is crucial. Physical examination, chest x-ray, ECG, and ECHO are non-invasive PAH screening procedures. This study found pulmonary hypertension in COPD patients by assessing chest X-ray, electrocardiographic, and echocardiographic findings and their correlation with ECG alterations. The study followed the cross-sectional observational study, in the national academy of medical sciences (NAMS), Bir Hospital, Kathmandu-Nepal. Patients who visited OPD or admitted with signs and symptoms of COPD diagnosed clinically, radiologically and/or by pulmonary function test (PFT), both male and female, in Bir Hospital. Each individual was chosen by simple random sampling method. N=100 cases were selected randomly by lottery over a period of one year. Patient’s chart, patients’s OPD card or structural questionnaire were used as the research instrument where Patients were selected for data collection by simple random sampling method like lottery among those who visited OPD or admitted in ward of Bir hospital with diagnosis of COPD by history, physical and radiological examination and/or PFT criteria. The highest COPD prevalence was in people 70 or older (42%), with 66% of these being women. The most common risk factor for COPD was smoking (75%). Chest x-ray abnormalities most often were significant pulmonary conus (76%) and cardiomegaly (63%).The majority of ECG abnormalities in COPD patients with pulmonary artery hypertension were P-pulmonale (84%), R/S ratio <1 in V5 or V6, right axis deviation (49%), and right bundle branch block (41%).83% of patients met RVH ECG criteria. Hence it concluded that ECHO is better than ECG and chest x-ray at detecting pulmonary hypertension in COPD, but ECG has good specificity and sensitivity for RVH and pulmonary hypertension, so chest x-ray and ECG should be routine testing.

The respiratory sound classification paper focuses on analyzing sounds collected through the stethoscope at various chest positions to understand critical concerns of human breathing. Most doctors used a stethoscope for auscultation of lung sounds, along with heart sound, as a basic checkup and utilized later in predicting chest pathology. Using deep learning-based sound classification can build intelligent digital stethoscopes or understand overall health conditions later used along with other Lung organ analyses. The solution enables affordable and practical tools for remote parts of India, where doctors or specialists are not keen to work. Such solutions can be used as first analysis before further investigation by traveling to far-off specialist hospitals. The lung sound was prepared and hosted by a team from the University of Coimbra, Portugal; the University de Aveiro, Portugal; and the Aristotle University of Thessaloniki, Greece and the collaboration at Kaggle website lung sounds of patients who are Normal lung functioning and with some anomaly collected. The collected lung sounds are labeled as Normal, seven classes - Abnormal. Overall, eight levels are considered to classify using a deep learning model. The paper proposes a multipath convolution neural network (CNN) model for analyzing sound after exploring all approaches, like Recurrent Neural network (RNN), Machine learning models like SVM, and other CNN models as per paper [3]. Lung sound is processed using signal processing methods and converts wav files to Mel-frequency cepstral coefficients (MFCCs), Mel spectrogram, and chromatogram from a waveform or power spectrogram. Instead of using one feature among three, the paper built a Multipath CNN model to feed all three components to model and concatenate feature maps to classify the classes. The model was evaluated with various parameters like accuracy and F1 score. The proposed model gives 96.17% for the dataset used. The result implies data, and a multipath approach helps to fuse features extracted from different images and allows models to classify well.

Chronic respiratory diseases are among the most important causes of morbidity and mortality worldwide. According to the Global Burden of Disease Study 2017, chronic obstructive pulmonary disease (COPD) was the third leading cause of global disability-adjusted life-years (DALY) among non-communicable diseases, with an increase of 13.5% over the last decade.1 Asthma is associated with a high burden of disease and substantial economic and societal impact.2 Interstitial lung diseases (ILD) represent a heterogeneous group of chronic and often progressive lung conditions, with limited response to pharmacotherapy, increasing disability and incremental healthcare utilization.3 Despite advances in pharmacological treatments for these and other chronic respiratory diseases, a large proportion of patients remain symptomatic and suffer from frequent exacerbations and hospitalizations, and the course of these diseases cannot be reversed. In addition, chronic respiratory diseases are characterized by extra-pulmonary features, and many patients suffer from comorbidities.4-6 In addition to smoking, physical inactivity is a well-recognized disadvantageous lifestyle factor contributing to disease progression, impaired health status and exercise capacity and increased mortality in COPD and other chronic respiratory diseases.7, 8 Unhealthy diet, extremes in body weight and low muscle mass are associated with poor outcomes in patients with COPD and ILD.9, 10 Many patients also experience difficulties coping with their disease11 and have limited knowledge about their condition and poor self-management skills.12 Thus, there is a clear clinical indication for additional, comprehensive interventions optimizing the physical and psychological condition of these patients and inducing a healthier lifestyle. Pulmonary rehabilitation (PR) offers this potential. It takes the factors contributing to the current and future burden of disease in individual patients into account. Clinically relevant improvements in health status, exercise capacity and other outcomes have been convincingly demonstrated in patients with stable disease13 and following exacerbation.14 Consequently, PR is considered an essential component of respiratory disease management and is recommended by various international guidelines.15-17 Despite the abundance of literature reporting the benefits of PR for patients with chronic respiratory diseases, this intervention is faced with numerous challenges. A very limited proportion of potential PR candidates are referred for these programmes, and adherence is low. Lack of awareness regarding access and benefits and insufficient funding and availability of PR programmes are important issues related to this underuse.18 Motivated by the lack of resources and poor uptake of traditional centre-based PR, alternative models, including home-based PR with minimal resources, have been studied.19, 20 The clinical outcomes of these programmes were positive, but they have instigated an international debate about the definition of PR and its minimal core components. Programmes consisting of exercise training only may improve physical functioning but may downsize the complexity of chronic respiratory diseases into a single dimension. Therefore, such programmes neglect the need for an integrated approach of all treatable components of these conditions. To move ahead and avoid devaluation of the term PR, international experts must come to agreement on essential organizational aspects and components of comprehensive PR or introduce alternative terminology for other approaches. Another challenge is related to non-response to PR in a proportion of patients and the current lack of ability to select the appropriate patients for PR programmes.21 In the realm of ‘precision medicine’, researchers in PR need to focus on these non-responders, identify treatable traits and determine which components of PR will be effective and use the outcomes to innovate and increase the efficacy of PR. For this review series on rehabilitation for chronic lung diseases, we have invited a group of global experts in PR to highlight the opportunities and challenges of this field and to provide the readers of this journal with a state-of-the-art update on the organization and content of PR in various respiratory conditions. With this, we aim to increase the awareness, implementation, quality, customization and outcomes of PR in different states (stable or exacerbation) and stages of severity of these diseases. Our experts will discuss the urgency of conceptual clarity and international consensus on the term PR and its essential components. They will also focus on the need for tailored programmes that meet the needs of individual patients. In addition to achieving a consensus on meaningful and sustainable outcomes, opportunities and limitations of PR programmes in different settings will be discussed and balanced against availability and costs. The invited authors will also provide expertise on selection of patients for PR and the components of an integrated assessment. Readers will be educated on the personalized exercise training and educational components of PR, along with other interventions. The effects of PR on patients with COPD and on patients with non-COPD diagnoses, including ILD, bronchiectasis, lung cancer and pulmonary hypertension, will be presented. In addition, invited authors will focus on the effects and safety of PR during acute exacerbations or recovery from these events. The experts will also describe the latest developments in PR for the most severe patients, for example by combining exercise training with (high-flow) oxygen supplementation and non-invasive ventilation. Finally, they will show us the evidence regarding maintenance programmes and inform readers of the barriers and facilitators for long-term benefits of PR. PR is a powerful intervention for patients with chronic respiratory diseases. Whilst these diseases remain incurable, PR will help support patients to live their life to the max. We truly hope that this review series in Respirology will be informative and interesting and that it will increase awareness among healthcare professionals, policymakers, payers and patients regarding the role of PR in respiratory health. We also hope that these reviews will stimulate the implementation of PR in regions of the world where it is not available and encourage scientists to explore the uncertainties in PR and ultimately advance this field.

Objective To investigate the value of echocardiography in assessing the right heart function of patients with chronic obstructive pulmonary disease (COPD). Methods Forty-four COPD patients who were treated in Beijing Anzhen Hospital of Capital Medical University, from April 2016 to April 2017, were selected as COPD group; and 12 healthy physical examiners were included in healthy control group during the same period. Patients were divided into COPD with pulmonary hypertension (PH) group and COPD without PH group. All subjects were routinely examined by transthoracic echocardiography. The parameters of right heart function of all subjects were measured by echocardiography according to 2010 guideline of American Society of Echocardiography (ASE). Independent sample t test was used to compare echocardiographic routine parameters and recommended parameters of ASE guideline between COPD group and healthy control group. One-way analysis of variance was used to compare the routine parameters of echocardiography and the recommended parameters of the ASE guide in the patients of COPD with PH group and COPD without PH group and the healthy control group. SNK-q test was used for comparison between groups. Results The right ventricle diameter (RVD) in group COPD was wider than that in healthy control group [(20.68±4.21) mm vs (18.17±1.75) mm], and the difference was statistically significant (t=2.92, P=0.005). There was no significant difference in the right ventricular outflow tract (RVOT), main pulmonary artery diameter (MPAD), left ventricular end-diastolic diameter (LVEDD), left ventricular end-systolic diameter (LVESD) and left ventricular ejection fraction (LVEF) between the COPD group and the healthy control group. Compared with the healthy control group, the diameter of right ventricle basal segment in the COPD group increased [(35.92±8.12) mm vs (27.75±3.17) mm], tricuspid annular plane systolic excursion (TAPSE) decreased [(19.61±4.08) mm vs (22.67±2.67) mm], right ventricular index of myocardial performance (RIMP) increased [(0.52±0.10) cm/s vs (0.43±0.04) cm/s)], and the differences were statistically significant (t=3.39, P=0.001; t=-2.44, P=0.019; t=4.31, P<0.001). There was no significant difference in right atrium area, E/A, E/E' and S' between COPD group and healthy control group. There was no significant difference in RVOT, RVD, MPAD, LVEDD, LVESD and LVEF in the patients of COPD with PH group and COPD without PH group and in the healthy control group. There was no significant difference in the right atrium area, E/A, E/E', TAPSE and S'. The right ventricular basal segment diameter and RIMP of COPD with PH group and COPD without PH group were higher than those of healthy control group [(37.99±9.66) mm, (34.47±6.70) mm vs (27.75±3.17) mm; (0.54±0.13) cm/s, (0.51±0.08) cm/s vs (0.43±0.04) cm/s]. The differences were statistically significant (q=6.960, 4.905, 5.796, 4.348, all P<0.05). However, there was no significant difference in right ventricular basal segment diameter and RIMP between COPD with PH group and COPD without PH group. The RVWT of COPD with PH group was higher than that of COPD without PH group [(5.29±0.69) mm vs (4.54±0.70) mm], and the difference was statistically significant (t=3.313, P=0.002). Conclusions The method recommended in the ASE guidelines for this study was more sensitive than conventional methods for the detection of changes in the structure of the right heart. The change of the right ventricular structure was the first manifestation of right heart involvement in COPD patients, and then the systolic function of the right ventricle diminished. The long-term effect of pulmonary hypertension was thickening of the right ventricular wall at the early stage and then enlargement of the right ventricle. Key words: Echocardiography; Pulmonary disease, chronic obstructive; Ventricular function, right; Atrial function, right

Pulmonary rehabilitation utilization in patients with chronic respiratory diseases: 2014-2019.

PURPOSE: Dyspnea, fatigue, reduced exercise tolerance, peripheral muscle dysfunction and mood disorders are common features of many chronic respiratory disorders. The aim of this study was to investigate the differences for peripheral muscle strength and functional exercise capacity between the patients with chronic obstructive pulmonary disease (COPD) and pulmonary arterial hypertension (PAH). METHODS: Moderate to severe 14 COPD patients with mean age 66.64 ± 9.22 years old and New York Heart Association Class II-III 22 PAH patients with mean age 61.27 ± 8.76 years old a, totally 36 patients were evaluated for this study. Peripheral muscle strength was measured by hand-held dynamometer for quadriceps femoris muscle and grip strength was evaluated by handgrip. Functional exercise capacity was measured by six-minute walk test. Differences between the groups were analysed by two independent sample t test. RESULTS: The strength of qudriceps femoris muscle for dominant extremity was 19.35 ± 5. 81 kg in patients with PAH and 32.35 ± 6.10 kg in patients with COPD. The grip strength for dominant extremity was 23.57 9.64 ± kg in patients with PAH and 30.28 ± 8.73 kg in patients with COPD There was statistically significantly decrease in muscle strength for quadriceps femoris and grip strength in favour of PAH patients between the groups (p <0.05). The functional exercise capacity was 392.72 ± 72.90 m as completing 74.4 % of predicted value in patients with PAH and 408.21 ± 70.59 m as completing 79.6 % of predicted value in patients with COPD. Although functional capacity was reduced both of the diseases there was no significantly difference between the groups (p > 0.05). CONCLUSIONS: Both of patients with COPD and PAH showed declined functional exercise capacity whereas muscle strength was significantly lower in PAH patients. This result indicated that severe COPD patients and PAH patients present similar functional exercise capacity while muscle strength was affected in PAH patients more than COPD patients.

Background: Chronic obstructive pulmonary disease (COPD) and pulmonary arterial hypertension (PAH) are common and underdiagnosed medical conditions in India. Prevalence of these chronic diseases is high both in rural and urban areas. However, the exact prevalence of PAH in Indian COPD patients is unclear. Comorbid conditions like PAH have a great impact on the outcome of COPD in the form of severity, exacerbations, morbidity, and mortality. Right heart catheterization remains the gold standard test for diagnosis of PAH, but it is invasive and practically not feasible. Aims and objectives: The present study objective was to screen COPD patients for PAH using two-dimensional transthoracic Doppler echocardiography (ECHO) in Tertiary Care Hospital. Results: A total of 2040 patients with a confirmed diagnosis of COPD were enrolled in the study. Among these patients, 1509 were males (73.9%), 531 were females (23.06%), and 1428 were known or ex-smokers (70%). None of the females were smokers, but there was a history of biomass fuel exposure for >10 years. As per global initiative for chronic obstructive lung disease criteria, mild, moderate, severe, and very severe COPD was noted in 525,629,511, and 375 patients, respectively. When they were screened using ECHO, prevalence of PAH was 41.96% as 856 subjects had PAH. Prevalence of PAH among mild, moderate, severe, and very severe COPD was 23.8%, 34.81%, 48.53%, and 70.4%, respectively. There was a linear relationship between PAH and severity of COPD. Conclusion: PAH is very common in COPD patients in India. As the severity of COPD increased, the frequency and degree of PAH also increased. ECHO is an excellent tool for detection of PAH in COPD patients. All patients with severe to very severe COPD should be routinely screened by ECHO for PAH and with mild to moderate COPD, who have dyspnea out of proportion to their clinical condition should also be screened for PAH.

We designed a computer-based respiratory sound analysis system to identify pediatric normal lung sound. To verify the validity of the computer-based respiratory sound analysis system. First we downloaded the standard lung sounds from the network database (website: http: //www.easyauscultation.com/lung-sounds-reference-guide) and recorded 3 samples of abnormal loud sound (rhonchi, wheeze and crackles) from three patients of The Department of Pediatrics, the First Affiliated Hospital of Xiamen University. We regarded such lung sounds as"reference lung sounds". The"test lung sounds"were recorded from 29 children form Kindergarten of Xiamen University. we recorded lung sound by portable electronic stethoscope and valid lung sounds were selected by manual identification. We introduced Mel-frequency cepstral coefficient (MFCC) to extract lung sound features and dynamic time warping (DTW) for signal classification. We had 39 standard lung sounds, recorded 58 test lung sounds. This computer-based respiratory sound analysis system was carried out in 58 lung sound recognition, correct identification of 52 times, error identification 6 times. Accuracy was 89.7%. Based on MFCC and DTW, our computer-based respiratory sound analysis system can effectively identify healthy lung sounds of children (accuracy can reach 89.7%), fully embodies the reliability of the lung sounds analysis system.

Chronic obstructive pulmonary disease significantly affects the quality of life, considerably limiting the physical capabilities of patients and is one of the main causes of morbidity and mortality in modern society. Currently, the problem of the comorbidity for the patients with chronic obstructive pulmonary disease is very relevant. The most common in patients of this category is hypertension. Today, it has been established that disorders in the immune system are detected in a very large group of diseases, including both chronic obstructive pulmonary disease and hypertension. The purpose of the study was to evaluate the dynamics of inflammatory markers in the patients with pulmonary hypertension on the background of chronic obstructive pulmonary disease in combination with hypertension under the influence of treatment. Material and methods. The results of the study are based on data from a comprehensive survey of 170 patients aged 40 to 65 years with chronic obstructive pulmonary disease, of which 123 had pulmonary hypertension (of which 87 had stage II hypertension and 36 were without hypertension) and 47 people were without pulmonary hypertension. The patients were chosen in the period from 2015 to 2018 on the basis of the municipal institution "Zaporizhzhia regional clinical hospital" of the Zaporizhzhia regional council. Results and discussion. The results of this study indicate the role of inflammation, or rather the levels of hs-CRP and IL-6, in the pathogenesis of pulmonary hypertension in patients with chronic obstructive pulmonary disease. The level of hs-CRP among the patients with pulmonary hypertension on a background of chronic obstructive pulmonary disease made up 10.46 [6,24 ; 15,30] mg/l and was significantly higher as against the values in the group of patients with chronic obstructive pulmonary disease without pulmonary hypertension and compared with group of healthy persons. The increase in the level of IL-6 in the group of patients having pulmonary hypertension with chronic obstructive pulmonary disease is significantly higher by 57 % compared to the value in the group of chronic obstructive pulmonary disease without pulmonary hypertension and by 7.4 times the value in the group of healthy individuals (p &lt;0.05). Comparing the subgroups of patients, and depending on the variant of exacerbation of chronic obstructive pulmonary disease, the level of hs-CRP and IL-6 was considerably higher in the subgroup with the infectious type of exacerbation compared with the subgroup of the non-infectious type of exacerbation of chronic obstructive pulmonary disease. Conclusion. After 12 months of treatment, when roflumilast was added to the basic therapy, we revealed a statistically significant difference in the levels of IL-6, hs-CRP and IL-10 in the plasma of patients with pulmonary hypertension on the background of chronic obstructive pulmonary disease combined with hypertension

The aim of the study was to inspect the acoustic properties and sleep characteristics of a preapneic snoring sound. The feasibility of forecasting upcoming respiratory events by snoring sound was also investigated. Participants with habitual snoring or a heavy breathing sound during sleep were recruited consecutively. Polysomnography was conducted, and snoring-related breathing sound was recorded simultaneously. Acoustic features and sleep features were extracted from 30-second samples, and a machine learning algorithm was used to establish 2 prediction models. A total of 74 eligible participants were included. Model 1, tested by 5-fold cross-validation, achieved an accuracy of 0.92 and an area under the curve of 0.94 for respiratory event prediction. Model 2, with acoustic features and sleep information tested by Leave-One-Out cross-validation, had an accuracy of 0.78 and an area under the curve of 0.80. Sleep position was found to be the most important among all sleep features contributing to the performance of the 2 models. Preapneic sound presented unique acoustic characteristics, and snoring-related breathing sound could be deployed as a real-time apneic event predictor. The models, combined with sleep information, serve as a promising tool for an early warning system to forecast apneic events. Wang B, Yi X, Gao J, etal. Real-time prediction of upcoming respiratory events via machine learning using snoring sound signal. J Clin Sleep Med. 2021;17(9):1777-1784.