Predicting spirometry readings using cough sound features and regression

A noval pulmonary function evaluation method based on ResNet50 + SVR model and cough

Traditionally, the clinical evaluation of respiratory diseases was pulmonary function testing, which can be used for the detection of severity and prognosis through pulmonary function parameters. However, this method is limited by the complex process, which is impossible for patients to monitor daily. In order to evaluate pulmonary function parameters conveniently with less time and location restrictions, cough sound is the substitute parameter. In this paper, 371 cough sounds segments from 150 individuals were separated into 309 and 62 as the training and test samples. Short-time Fourier transform (STFT) was applied to transform cough sound into spectrogram, and ResNet50 model was used to extract 2048-dimensional features. Through support vector regression (SVR) model with biological attributes, the data were regressed with pulmonary function parameters, FEV1, FEV1%, FEV1/FVC, FVC, FVC%, and the performance of this models was evaluated with fivefold cross-validation. Combines with deep learning and machine learning technologies, the better results in the case of small samples were achieved. Using the coefficient of determination (R2), the ResNet50 + SVR model shows best performance in five basic pulmonary function parameters evaluation as FEV1(0.94), FEV1%(0.84), FEV1/FVC(0.68), FVC(0.92), and FVC%(0.72). This ResNet50 + SVR hybrid model shows excellent evaluation of pulmonary function parameters during coughing, making it possible to realize a simple and rapid evaluation for pneumonia patients. The technology implemented in this paper is beneficial in judge the patient's condition, realize early screening of respiratory diseases, evaluate postoperative disease changes and detect respiratory infectious diseases without time and location restrictions.

A smartphone-based algorithm comprising cough analysis and patient-reported symptoms identifies acute exacerbations of asthma: a prospective, double blind, diagnostic accuracy study

Objective: Early and accurate recognition of asthma exacerbations reduces the duration and risk of hospitalization. Current diagnostic methods depend upon patient recognition of symptoms, expert clinical examination, or measures of lung function. Here, we aimed to develop and test the accuracy of a smartphone-based diagnostic algorithm that analyses five cough events and five patient-reported features (age, fever, acute or productive cough and wheeze) to detect asthma exacerbations.Methods: We conducted a double-blind, prospective, diagnostic accuracy study comparing the algorithm with expert clinical opinion and formal lung function testing. Results: One hundred nineteen participants >12 years with a physician-diagnosed history of asthma were recruited from a hospital in Perth, Western Australia: 46 with clinically confirmed asthma exacerbations, 73 with controlled asthma. The groups were similar in median age (54yr versus 60yr, p=0.72) and sex (female 76% versus 70%, p=0.5). The algorithm's positive percent agreement (PPA) with the expert clinical diagnosis of asthma exacerbations was 89% [95% CI: 76%, 96%]. The negative percent agreement (NPA) was 84% [95% CI: 73%, 91%]. The algorithm's performance for asthma exacerbations diagnosis exceeded its performance as a detector of patient-reported wheeze (sensitivity, 63.7%). Patient-reported wheeze in isolation was an insensitive marker of asthma exacerbations (PPA=53.8%, NPA=49%). Conclusions: Our diagnostic algorithm accurately detected the presence of an asthma exacerbation as a point-of-care test without requiring clinical examination or lung function testing. This method could improve the accuracy of telehealth consultations and might be helpful in Asthma Action Plans and patient-initiated therapy.

AI-enabled remote monitoring of vital signs for COVID-19: methods, prospects and challenges

The COVID-19 pandemic has overwhelmed the existing healthcare infrastructure in many parts of the world. Healthcare professionals are not only over-burdened but also at a high risk of nosocomial transmission from COVID-19 patients. Screening and monitoring the health of a large number of susceptible or infected individuals is a challenging task. Although professional medical attention and hospitalization are necessary for high-risk COVID-19 patients, home isolation is an effective strategy for low and medium risk patients as well as for those who are at risk of infection and have been quarantined. However, this necessitates effective techniques for remotely monitoring the patients’ symptoms. Recent advances in Machine Learning (ML) and Deep Learning (DL) have strengthened the power of imaging techniques and can be used to remotely perform several tasks that previously required the physical presence of a medical professional. In this work, we study the prospects of vital signs monitoring for COVID-19 infected as well as quarantined individuals by using DL and image/signal-processing techniques, many of which can be deployed using simple cameras and sensors available on a smartphone or a personal computer, without the need of specialized equipment. We demonstrate the potential of ML-enabled workflows for several vital signs such as heart and respiratory rates, cough, blood pressure, and oxygen saturation. We also discuss the challenges involved in implementing ML-enabled techniques.

A forced cough sound based pulmonary function assessment method by using machine learning

Pulmonary function testing (PFT) has important clinical value for the early detection of lung diseases, assessment of the disease severity, causes identification of dyspnea, and monitoring of critical patients. However, traditional PFT can only be carried out in a hospital environment, and it is challenging to meet the needs for daily and frequent evaluation of chronic respiratory diseases. In this study, we propose a novel method for accurately assessing pulmonary function by analyzing recorded forced cough sounds by mobile device without time and location restrictions. In the experiment, 309 clips of cough sound segments were separated from 133 patients who underwent PFT by using Audacity software. There are 247 clips of training samples and 62 clips of testing samples. Totally 52 features were extracted from the dataset, and principal component analysis (PCA) was used for feature reduction. Combined with biological attributes, the normalized features were regressed by using machine learning models with pulmonary function parameters (i.e., FEV1, FVC, FEV1/FVC, FEV1%, and FVC%). And a 5-fold cross-validation was applied to evaluate the performance of the regression models. As described in the experimental result, the result of coefficient of determination (R2) indicates that the support vector regression (SVR) model performed best in assessing FVC (0.84), FEV1% (0.61), and FVC% (0.62) among these models. The gradient boosting regression (GBR) model performs best in evaluating FEV1 (0.86) and FEV1/FVC (0.54). The result confirmed that the proposed method was capable of accurately assessing pulmonary function with forced cough sound. Besides, the cough sound sampling by a smartphone made it possible to conduct sampling and assess pulmonary function frequently in the home environment.

The Acoustic Dissection of Cough: Diving Into Machine Listening-based COVID-19 Analysis and Detection

OBJECTIVES

The coronavirus disease 2019 (COVID-19) has caused a crisis worldwide. Amounts of efforts have been made to prevent and control COVID-19's transmission, from early screenings to vaccinations and treatments. Recently, due to the spring up of many automatic disease recognition applications based on machine listening techniques, it would be fast and cheap to detect COVID-19 from recordings of cough, a key symptom of COVID-19. To date, knowledge of the acoustic characteristics of COVID-19 cough sounds is limited but would be essential for structuring effective and robust machine learning models. The present study aims to explore acoustic features for distinguishing COVID-19 positive individuals from COVID-19 negative ones based on their cough sounds.

METHODS

By applying conventional inferential statistics, we analyze the acoustic correlates of COVID-19 cough sounds based on the ComParE feature set, i.e., a standardized set of 6,373 acoustic higher-level features. Furthermore, we train automatic COVID-19 detection models with machine learning methods and explore the latent features by evaluating the contribution of all features to the COVID-19 status predictions.

RESULTS

The experimental results demonstrate that a set of acoustic parameters of cough sounds, e.g., statistical functionals of the root mean square energy and Mel-frequency cepstral coefficients, bear essential acoustic information in terms of effect sizes for the differentiation between COVID-19 positive and COVID-19 negative cough samples. Our general automatic COVID-19 detection model performs significantly above chance level, i.e., at an unweighted average recall (UAR) of 0.632, on a data set consisting of 1,411 cough samples (COVID-19 positive/negative: 210/1,201).

CONCLUSIONS

Based on the acoustic correlates analysis on the ComParE feature set and the feature analysis in the effective COVID-19 detection approach, we find that several acoustic features that show higher effects in conventional group difference testing are also higher weighted in the machine learning models.

Prediction of Pulmonary Function Parameters Based on a Combination Algorithm

Objective: Pulmonary function parameters play a pivotal role in the assessment of respiratory diseases. However, the accuracy of the existing methods for the prediction of pulmonary function parameters is low. This study proposes a combination algorithm to improve the accuracy of pulmonary function parameter prediction. Methods: We first established a system to collect volumetric capnography and then processed the data with a combination algorithm to predict pulmonary function parameters. The algorithm consists of three main parts: a medical feature regression structure consisting of support vector machines (SVM) and extreme gradient boosting (XGBoost) algorithms, a sequence feature regression structure consisting of one-dimensional convolutional neural network (1D-CNN), and an error correction structure using improved K-nearest neighbor (KNN) algorithm. Results: The root mean square error (RMSE) of the pulmonary function parameters predicted by the combination algorithm was less than 0.39L and the R² was found to be greater than 0.85 through a ten-fold cross-validation experiment. Conclusion: Compared with the existing methods for predicting pulmonary function parameters, the present algorithm can achieve a higher accuracy rate. At the same time, this algorithm uses specific processing structures for different features, and the interpretability of the algorithm is ensured while mining the feature depth information.

Predicting Pulmonary Function From the Analysis of Voice: A Machine Learning Approach

Introduction

To self-monitor asthma symptoms, existing methods (e.g. peak flow metre, smart spirometer) require special equipment and are not always used by the patients. Voice recording has the potential to generate surrogate measures of lung function and this study aims to apply machine learning approaches to predict lung function and severity of abnormal lung function from recorded voice for asthma patients.

Methods

A threshold-based mechanism was designed to separate speech and breathing from 323 recordings. Features extracted from these were combined with biological factors to predict lung function. Three predictive models were developed using Random Forest (RF), Support Vector Machine (SVM), and linear regression algorithms: (a) regression models to predict lung function, (b) multi-class classification models to predict severity of lung function abnormality, and (c) binary classification models to predict lung function abnormality. Training and test samples were separated (70%:30%, using balanced portioning), features were normalised, 10-fold cross-validation was used and model performances were evaluated on the test samples.

Results

The RF-based regression model performed better with the lowest root mean square error of 10·86. To predict severity of lung function impairment, the SVM-based model performed best in multi-class classification (accuracy = 73.20%), whereas the RF-based model performed best in binary classification models for predicting abnormal lung function (accuracy = 85%).

Conclusion

Our machine learning approaches can predict lung function, from recorded voice files, better than published approaches. This technique could be used to develop future telehealth solutions including smartphone-based applications which have potential to aid decision making and self-monitoring in asthma.

Cough Sounds Recorded via Smart Devices as Useful Non-Invasive Digital Biomarkers of Aspiration Risk: A Case Report

Spirometer measurements can reflect cough strength but might not be routinely available for patients with severe neurological or medical conditions. A digital device that can record and help track abnormal cough sound changes serially in a noninvasive but reliable manner would be beneficial for monitoring such individuals. This report includes two cases of respiratory distress whose cough changes were monitored via assessments performed using recordings made with a digital device. The cough sounds were recorded using an iPad (Apple, Cupertino, CA, USA) through an embedded microphone. Cough sounds were recorded at the bedside, with no additional special equipment. The two patients were able to complete the recordings with no complications. The maximum root mean square values obtained from the cough sounds were significantly reduced when both cases were diagnosed with aspiration pneumonia. In contrast, higher values became apparent when the patients demonstrated a less severe status. Based on an analysis of our two cases, the patients’ cough sounds recorded with a commercial digital device show promise as potential digital biomarkers that may reflect aspiration risk related to attenuated cough force. Serial monitoring aided the decision making to resume oral feeding. Future studies should further explore the clinical utility of this technique.

Identifying acute exacerbations of chronic obstructive pulmonary disease using patient-reported symptoms and cough feature analysis

Acute exacerbations of chronic obstructive pulmonary disease (AECOPD) are commonly encountered in the primary care setting, though the accurate and timely diagnosis is problematic. Using technology like that employed in speech recognition technology, we developed a smartphone-based algorithm for rapid and accurate diagnosis of AECOPD. The algorithm incorporates patient-reported features (age, fever, and new cough), audio data from five coughs and can be deployed by novice users. We compared the accuracy of the algorithm to expert clinical assessment. In patients with known COPD, the algorithm correctly identified the presence of AECOPD in 82.6% (95% CI: 72.9–89.9%) of subjects (n = 86). The absence of AECOPD was correctly identified in 91.0% (95% CI: 82.4–96.3%) of individuals (n = 78). The diagnostic agreement was maintained in milder cases of AECOPD (PPA: 79.2%, 95% CI: 68.0–87.8%), who typically comprise the cohort presenting to primary care. The algorithm may aid early identification of AECOPD and be incorporated in patient self-management plans.

The diagnosis of respiratory disease in children using a phone-based cough and symptom analysis algorithm: The smartphone recordings of cough sounds 2 (SMARTCOUGH-C 2) trial design

The diagnosis of acute respiratory diseases in children can be challenging, and no single objective diagnostic test exists for common pediatric respiratory diseases. Previous research has demonstrated that ResAppDx, a cough sound and symptom-based analysis algorithm, can identify common respiratory diseases at the point of care. We present the study protocol for SMARTCOUGH-C 2, a prospective diagnostic accuracy trial of a cough and symptom-based algorithm in a cohort of children presenting with acute respiratory diseases. The objective of the study is to assess the performance characteristics of the ResAppDx algorithm in the diagnosis of common pediatric acute respiratory diseases.

Stratifying asthma severity in children using cough sound analytic technology

Introduction: Asthma is a common childhood respiratory disorder characterized by wheeze, cough and respiratory distress responsive to bronchodilator therapy. Asthma severity can be determined by subjective, manual scoring systems such as the Pulmonary Score (PS). These systems require significant medical training and expertise to rate clinical findings such as wheeze characteristics, and work of breathing. In this study, we report the development of an objective method of assessing acute asthma severity based on the automated analysis of cough sounds.Methods: We collected a cough sound dataset from 224 children; 103 without acute asthma and 121 with acute asthma. Using this database coupled with clinical diagnoses and PS determined by a clinical panel, we developed a machine classifier algorithm to characterize the severity of airway constriction. The performance of our algorithm was then evaluated against the PS from a separate set of patients, independent of the training set.Results: The cough-only model discriminated no/mild disease (PS 0-1) from severe disease (PS 5,6) but required a modified respiratory rate calculation to separate very severe disease (PS > 6). Asymptomatic children (PS 0) were separated from moderate asthma (PS 2-4) by the cough-only model without the need for clinical inputs.Conclusions: The PS provides information in managing childhood asthma but is not readily usable by non-medical personnel. Our method offers an objective measurement of asthma severity which does not rely on clinician-dependent inputs. It holds potential for use in clinical settings including improving the performance of existing asthma-rating scales and in community-management programs.AbbreviationsAMaccessory muscleBIbreathing indexCIconfidence intervalFEV1forced expiratory volume in one secondLRlogistic regressionPEFRpeak expiratory flow ratePSpulmonary scoreRRrespiratory rateSDstandard deviationSEstandard errorWAWestern Australia.

A prospective multicentre study testing the diagnostic accuracy of an automated cough sound centred analytic system for the identification of common respiratory disorders in children

BACKGROUND

The differential diagnosis of paediatric respiratory conditions is difficult and suboptimal. Existing diagnostic algorithms are associated with significant error rates, resulting in misdiagnoses, inappropriate use of antibiotics and unacceptable morbidity and mortality. Recent advances in acoustic engineering and artificial intelligence have shown promise in the identification of respiratory conditions based on sound analysis, reducing dependence on diagnostic support services and clinical expertise. We present the results of a diagnostic accuracy study for paediatric respiratory disease using an automated cough-sound analyser.

METHODS

We recorded cough sounds in typical clinical environments and the first five coughs were used in analyses. Analyses were performed using cough data and up to five-symptom input derived from patient/parent-reported history. Comparison was made between the automated cough analyser diagnoses and consensus clinical diagnoses reached by a panel of paediatricians after review of hospital charts and all available investigations.

RESULTS

A total of 585 subjects aged 29 days to 12 years were included for analysis. The Positive Percent and Negative Percent Agreement values between the automated analyser and the clinical reference were as follows: asthma (97, 91%); pneumonia (87, 85%); lower respiratory tract disease (83, 82%); croup (85, 82%); bronchiolitis (84, 81%).

CONCLUSION

The results indicate that this technology has a role as a high-level diagnostic aid in the assessment of common childhood respiratory disorders.

TRIAL REGISTRATION

Australian and New Zealand Clinical Trial Registry (retrospective) - ACTRN12618001521213 : 11.09.2018.

Falls rate increase and foot dorsal flexion limitations are exhibited in patients who suffer from asthma: A novel case-control study

Purpose: Based on the possible association between reduced foot dorsiflexion and high risk of falls, the main objective was to determine the ankle and 1º metatarsophalangeal joint (1stMTTP) dorsiflexion range of motion and falls rate in patients with asthma compared to healthy matched-paired controls. Methods: A case-control study was carried out. Eighty participants were recruited and divided into patients with asthma (case group; n=40) and matched-paired healthy participants (control group; n=40). Foot dorsal flexion range of motion (assessed by the Weight-Bearing Lunge Test [WBLT]) and falls rate (evaluated as falls number during the prior year) were considered as the primary outcomes. Indeed, ankle dorsiflexion was measured by a mobile app (º) and a tape measure (cm) as well as 1stMTTP dorsiflexion was determined by and universal goniometer (º). Results: Statistically significant differences (P<.05) showed that patients with asthma presented a greater falls rate than healthy participants and reduced bilateral ankle and 1stMTTP dorsiflexion ranges of motion than healthy participants, except for the left ankle dorsiflexion measured as degrees (P>.05). Conclusions: These study findings showed that a falls rate increase and bilateral foot dorsal flexion limitations of the ankle and 1stMTTP joints are exhibited in patients who suffer from asthma.

Predicting forced vital capacity (FVC) using support vector regression (SVR)

OBJECTIVE

Spirometry, as the gold standard approach in the diagnosis of chronic obstructive pulmonary disease (COPD), has strict end of test (EOT) criteria (e.g. complete exhalation), which cannot be met by patients with compromised health states. Thus, significant parameters measured by spirometry, such as forced vital capacity (FVC), have limited accuracies. To address this issue, the present study aimed to develop models based on support vector regression (SVR) to predict values of FVC under the condition that the EOT criteria were not fully met.

APPROACH

The prediction models for the quantification of FVC were developed based on SVR. A total of 354 subjects underwent conventional spirometry (CS), and the resulting data of forced expiratory volumes in 1 s (FEV1), peak expiratory flow (PEF), age and gender were used as input features, while the resulting values of the FVC were used as the target feature in the prediction models. Next, three prediction models (mixed model, normal model and abnormal model) were established according to the criterion in the diagnosis of COPD that a postbronchodilator shows an FEV1/FVC ratio lower than 0.70. Then, 35 subjects were recruited to be tested using both CS and a low-degree-of-EOT criteria spirometry (LDCS), which did not fully meet the EOT criteria of CS. In LDCS, subjects were allowed to terminate the procedure at their own will at any time after the technicians had assumed that both acceptable values of FEV1 and PEF had been obtained. Quantified values of FVC derived from both CS and LDCS were compared to validate the performances of the developed prediction models.

MAIN RESULTS

The FVC prediction performances of the normal model and abnormal model were better than that of the mixed model. The root mean squared error are lower than 0.35 l and the accuracies are higher up to 95%. One-tailed t test results demonstrate that the absolute differences in the measured and predicted values are not significantly different from 0.15 l for both the abnormal model and the normal model.

SIGNIFICANCE

Our study shows the possibility of predicting FVC with acceptable precision in cases where the EOT criteria of spirometry were not fully met, which can be beneficial for patients who cannot or did not achieve full exhalation in spirometry.

Tarsal Tunnel Mechanosensitivity Is Increased in Patients with Asthma: A Case-Control Study

Background: Based on changes in lung function and musculoskeletal disorders in patients with asthma, this study aimed to compare the tarsal tunnel and fibular bone pressure pain thresholds (PPTs) of patients with asthma and healthy matched-paired controls. Methods: A case-control study was performed. One hundred participants were recruited: 50 asthma patients and 50 healthy matched-paired controls. Bilaterally, tarsal tunnel and fibula bone PPTs were registered. Results: Statistically significant differences (p < 0.01) were shown bilaterally for tarsal tunnel PPT. With the exception of fibula PPT (p > 0.05), asthma patients presented less tarsal tunnel PPT than healthy participants. Statistically significant differences (p < 0.05) were shown for two linear regression prediction models of the right (R² = 0.279) and left (R² = 0.249) tarsal tunnels PPTs as dependent variables, and based on sex, group, contralateral tarsal tunnel PPT and ipsilateral fibula PPT as independent variables. Conclusions: The study findings showed that a bilateral tarsal tunnel mechanosensitivity increase is exhibited in patients diagnosed with asthma. The presence of asthma may bilaterally predict the PPT of tarsal tunnel. These findings may suggest the presence of central sensitization in asthma patients, which could clinically predispose them to musculoskeletal disorders, such as tarsal tunnel syndrome.