Detection of sputum by interpreting the time-frequency distribution of respiratory sound signal using image processing techniques

RespirAnalyzer: an R package for analyzing data from continuous monitoring of respiratory signals

Motivation

The analysis of data obtained from continuous monitoring of respiratory signals (CMRS) holds significant importance in improving patient care, optimizing sports performance, and advancing scientific understanding in the field of respiratory health.

Results

The R package RespirAnalyzer provides an analytic tool specifically for feature extraction, fractal and complexity analysis for CMRS data. The package covers a wide and comprehensive range of data analysis methods including obtaining inter-breath intervals (IBI) series, plotting time series, obtaining summary statistics of IBI series, conducting power spectral density, multifractal detrended fluctuation analysis (MFDFA) and multiscale sample entropy analysis, fitting the MFDFA results with the extended binomial multifractal model, displaying results using various plots, etc. This package has been developed from our work in directly analyzing CMRS data and is anticipated to assist fellow researchers in computing the related features of their CMRS data, enabling them to delve into the clinical significance inherent in these features.

Availability and implementation

The package for Windows is available from both Comprehensive R Archive Network (CRAN): https://cran.r-project.org/web/packages/RespirAnalyzer/index.html and GitHub: https://github.com/dongxinzheng/RespirAnalyzer.

Breathing sounds analysis system for early detection of airway problems in patients with a tracheostomy tube

To prevent immediate mortality in patients with a tracheostomy tube, it is essential to ensure timely suctioning or replacement of the tube. Breathing sounds at the entrance of tracheostomy tubes were recorded with a microphone and analyzed using a spectrogram to detect airway problems. The sounds were classified into three categories based on the waveform of the spectrogram according to the obstacle status: normal breathing sounds (NS), vibrant breathing sounds (VS) caused by movable obstacles, and sharp breathing sounds (SS) caused by fixed obstacles. A total of 3950 breathing sounds from 23 patients were analyzed. Despite neither the patients nor the medical staff recognizing any airway problems, the number and percentage of NS, VS, and SS were 1449 (36.7%), 1313 (33.2%), and 1188 (30.1%), respectively. Artificial intelligence (AI) was utilized to automatically classify breathing sounds. MobileNet and Inception_v3 exhibited the highest sensitivity and specificity scores of 0.9441 and 0.9414, respectively. When classifying into three categories, ResNet_50 showed the highest accuracy of 0.9027, and AlexNet showed the highest accuracy of 0.9660 in abnormal sounds. Classifying breathing sounds into three categories is very useful in deciding whether to suction or change the tracheostomy tubes, and AI can accomplish this with high accuracy.

A Wearable Multimodal Wireless Sensing System for Respiratory Monitoring and Analysis

Wireless sensing systems are required for continuous health monitoring and data collection. It allows for patient data collection in real time rather than through time-consuming and expensive hospital or lab visits. This technology employs wearable sensors, signal processing, and wireless data transfer to remotely monitor patients' health. The research offers a novel approach to providing primary diagnostics remotely with a digital health system for monitoring pulmonary health status using a multimodal wireless sensor device. The technology uses a compact wearable with new integration of acoustics and biopotentials sensors to monitor cardiovascular and respiratory activity to provide comprehensive and fast health status monitoring. Furthermore, the small wearable sensor size may stick to human skin and record heart and lung activities to monitor respiratory health. This paper proposes a sensor data fusion method of lung sounds and cardiograms for potential real-time respiration pattern diagnostics, including respiratory episodes like low tidal volume and coughing. With a p-value of 0.003 for sound signals and 0.004 for electrocardiogram (ECG), preliminary tests demonstrated that it was possible to detect shallow breathing and coughing at a meaningful level.

Acoustic Lung Imaging Utilized in Continual Assessment of Patients with Obstructed Airway: A Systematic Review

Smart respiratory therapy is enabled by continual assessment of lung functions. This systematic review provides an overview of the suitability of equipment-to-patient acoustic imaging in continual assessment of lung conditions. The literature search was conducted using Scopus, PubMed, ScienceDirect, Web of Science, SciELO Preprints, and Google Scholar. Fifteen studies remained for additional examination after the screening process. Two imaging modalities, lung ultrasound (LUS) and vibration imaging response (VRI), were identified. The most common outcome obtained from eleven studies was positive observations of changes to the geographical lung area, sound energy, or both, while positive observation of lung consolidation was reported in the remaining four studies. Two different modalities of lung assessment were used in eight studies, with one study comparing VRI against chest X-ray, one study comparing VRI with LUS, two studies comparing LUS to chest X-ray, and four studies comparing LUS in contrast to computed tomography. Our findings indicate that the acoustic imaging approach could assess and provide regional information on lung function. No technology has been shown to be better than another for measuring obstructed airways; hence, more research is required on acoustic imaging in detecting obstructed airways regionally in the application of enabling smart therapy.

A wearable device for at-home obstructive sleep apnea assessment: State-of-the-art and research challenges

In the last 3 years, almost all medical resources have been reserved for the screening and treatment of patients with coronavirus disease (COVID-19). Due to a shortage of medical staff and equipment, diagnosing sleep disorders, such as obstructive sleep apnea (OSA), has become more difficult than ever. In addition to being diagnosed using polysomnography at a hospital, people seem to pay more attention to alternative at-home OSA detection solutions. This study aims to review state-of-the-art assessment techniques for out-of-center detection of the main characteristics of OSA, such as sleep, cardiovascular function, oxygen balance and consumption, sleep position, breathing effort, respiratory function, and audio, as well as recent progress in the implementation of data acquisition and processing and machine learning techniques that support early detection of severe OSA levels.

Quantification of respiratory sounds by a continuous monitoring system can be used to predict complications after extubation: a pilot study

To show that quantification of abnormal respiratory sounds by our developed device is useful for predicting respiratory failure and airway problems after extubation. A respiratory sound monitoring system was used to collect respiratory sounds in patients undergoing extubation. The recorded respiratory sounds were subsequently analyzed. We defined the composite poor outcome as requiring any of following medical interventions within 48 h as defined below. This composite outcome includes reintubation, surgical airway management, insertion of airway devices, unscheduled use of noninvasive ventilation or high-flow nasal cannula, unscheduled use of inhaled medications, suctioning of sputum by bronchoscopy and unscheduled imaging studies. The quantitative values (QV) for each abnormal respiratory sound and inspiratory sound volume were compared between composite outcome groups and non-outcome groups. Fifty-seven patients were included in this study. The composite outcome occurred in 18 patients. For neck sounds, the QVs of stridor and rhonchi were significantly higher in the outcome group vs the non-outcome group. For anterior thoracic sounds, the QVs of wheezes, rhonchi, and coarse crackles were significantly higher in the outcome group vs the non-outcome group. For bilateral lateral thoracic sounds, the QV of fine crackles was significantly higher in the outcome group vs the non-outcome group. Cervical inspiratory sounds volume (average of five breaths) immediately after extubation was significantly louder in the outcome group vs non-outcome group (63.3 dB vs 54.3 dB, respectively; p < 0.001). Quantification of abnormal respiratory sounds and respiratory volume may predict respiratory failure and airway problems after extubation.

Does Facemask Impact Diagnostic During Pulmonary Auscultation?

Facemasks have been widely used in hospitals, especially since the emergence of the coronavirus 2019 (COVID-19) pandemic, often severely affecting respiratory functions. Masks protect patients from contagious airborne transmission, and are thus more specifically important for chronic respiratory disease (CRD) patients. However, masks also increase air resistance and thus work of breathing, which may impact pulmonary auscultation and diagnostic acuity, the primary respiratory examination. This study is the first to assess the impact of facemasks on clinical auscultation diagnostic. Lung sounds from 29 patients were digitally recorded using an electronic stethoscope. For each patient, one recording was taken wearing a surgical mask and one without. Recorded signals were segmented in breath cycles using an autocorrelation algorithm. In total, 87 breath cycles were identified from sounds with mask, and 82 without mask. Time-frequency analysis of the signals was used to extract comparison features such as peak frequency, median frequency, band power, or spectral integration. All the features extracted in frequency content, its evolution, or power did not significantly differ between respiratory cycles with or without mask. This early stage study thus suggests minor impact on clinical diagnostic outcomes in pulmonary auscultation. However, further analysis is necessary such as on adventitious sounds characteristics differences with or without mask, to determine if facemask could lead to no discernible diagnostic outcome in clinical practice.

Quantum algorithm for quicker clinical prognostic analysis: an application and experimental study using CT scan images of COVID-19 patients

BACKGROUND

In medical diagnosis and clinical practice, diagnosing a disease early is crucial for accurate treatment, lessening the stress on the healthcare system. In medical imaging research, image processing techniques tend to be vital in analyzing and resolving diseases with a high degree of accuracy. This paper establishes a new image classification and segmentation method through simulation techniques, conducted over images of COVID-19 patients in India, introducing the use of Quantum Machine Learning (QML) in medical practice.

METHODS

This study establishes a prototype model for classifying COVID-19, comparing it with non-COVID pneumonia signals in Computed tomography (CT) images. The simulation work evaluates the usage of quantum machine learning algorithms, while assessing the efficacy for deep learning models for image classification problems, and thereby establishes performance quality that is required for improved prediction rate when dealing with complex clinical image data exhibiting high biases.

RESULTS

The study considers a novel algorithmic implementation leveraging quantum neural network (QNN). The proposed model outperformed the conventional deep learning models for specific classification task. The performance was evident because of the efficiency of quantum simulation and faster convergence property solving for an optimization problem for network training particularly for large-scale biased image classification task. The model run-time observed on quantum optimized hardware was 52 min, while on K80 GPU hardware it was 1 h 30 min for similar sample size. The simulation shows that QNN outperforms DNN, CNN, 2D CNN by more than 2.92% in gain in accuracy measure with an average recall of around 97.7%.

CONCLUSION

The results suggest that quantum neural networks outperform in COVID-19 traits' classification task, comparing to deep learning w.r.t model efficacy and training time. However, a further study needs to be conducted to evaluate implementation scenarios by integrating the model within medical devices.

Use of carbon nanotube sensor for detecting postoperative abnormal respiratory waveforms

Background

This feasibility study aimed to detect respiratory waveforms from thoracic movements and evaluate if postoperative complications could be predicted using a carbon nanotube sensor.

Methods

Fifty patients who underwent lung resection for lung tumors were enrolled. The lung monitoring system of the carbon nanotube sensor was placed on bilateral chest walls across the 6^th–9^th ribs to measure chest wall motion. We examined the respiratory waveform in relation to surgical findings, postoperative course, and complications using Hilbert transform and Fast Fourier Transform (FFT).

Results

Of 50 patients (37 males, 13 females), 22 were included in the normal lung function group and 28 were included in the low lung function group. The respiratory rate and waveform indicated a regular pattern in the normal lung function group and the respiratory rate could be detected. Conversely, irregular respiratory pattern was detected in 70% of patients in the low lung function group. There was no significant different overall envelope peak value between operated side and non-operated side (0.195±0.05 and 0.18±0.06). In contrast, there was significantly high peak value in the presence of postoperative complications (P<0.05). And there was a significantly higher peak value in air leakage presence than air leakage absence in operated side (P=0.045).

Conclusions

The present study confirmed the feasibility of the sensor. It is promising in visualizing the respiratory state and detecting respiratory changes postoperatively.

Extraction and Analysis of Respiratory Motion Using a Comprehensive Wearable Health Monitoring System

Respiratory activity is an important vital sign of life that can indicate health status. Diseases such as bronchitis, emphysema, pneumonia and coronavirus cause respiratory disorders that affect the respiratory systems. Typically, the diagnosis of these diseases is facilitated by pulmonary auscultation using a stethoscope. We present a new attempt to develop a lightweight, comprehensive wearable sensor system to monitor respiration using a multi-sensor approach. We employed new wearable sensor technology using a novel integration of acoustics and biopotentials to monitor various vital signs on two volunteers. In this study, a new method to monitor lung function, such as respiration rate and tidal volume, is presented using the multi-sensor approach. Using the new sensor, we obtained lung sound, electrocardiogram (ECG), and electromyogram (EMG) measurements at the external intercostal muscles (EIM) and at the diaphragm during breathing cycles with 500 mL, 625 mL, 750 mL, 875 mL, and 1000 mL tidal volume. The tidal volumes were controlled with a spirometer. The duration of each breathing cycle was 8 s and was timed using a metronome. For each of the different tidal volumes, the EMG data was plotted against time and the area under the curve (AUC) was calculated. The AUC calculated from EMG data obtained at the diaphragm and EIM represent the expansion of the diaphragm and EIM respectively. AUC obtained from EMG data collected at the diaphragm had a lower variance between samples per tidal volume compared to those monitored at the EIM. Using cubic spline interpolation, we built a model for computing tidal volume from EMG data at the diaphragm. Our findings show that the new sensor can be used to measure respiration rate and variations thereof and holds potential to estimate tidal lung volume from EMG measurements obtained from the diaphragm.

A Phlegm Stagnation Monitoring Based on VDS Algorithm

In the nonmedical sputum monitoring system, a practical solution for phlegm stagnation care of patients was proposed. Through the camera, the video images of patients' laryngeal area were obtained in real time. After processing and analysis on these video frame images, the throat movement area was found out. A three-frame differential method was used to detect the throat moving targets. Anomalies were identified according to the information of moving targets and the proposed algorithm. Warning on the abnormal situation can help nursing personnel to deal with sputum blocking problem more effectively. To monitor the patients' situation in real time, this paper proposed a VDS algorithm, which extracted the speed characteristics of moving objects and combined with the DTW algorithm and SVM algorithm for sequence image classification. Phlegm stagnation symptoms of patients were identified timely for further medical care. In order to evaluate the effectiveness, our method was compared with the DTW, SVM, CTM, and HMM methods. The experimental results showed that this method had a higher recognition rate and was more practical in a nonmedical monitoring system.

A Novel Method for Automatic Identification of Breathing State

Sputum deposition blocks the airways of patients and leads to blood oxygen desaturation. Medical staff must periodically check the breathing state of intubated patients. This process increases staff workload. In this paper, we describe a system designed to acquire respiratory sounds from intubated subjects, extract the audio features, and classify these sounds to detect the presence of sputum. Our method uses 13 features extracted from the time-frequency spectrum of the respiratory sounds. To test our system, 220 respiratory sound samples were collected. Half of the samples were collected from patients with sputum present, and the remainder were collected from patients with no sputum present. Testing was performed based on ten-fold cross-validation. In the ten-fold cross-validation experiment, the logistic classifier identified breath sounds with sputum present with a sensitivity of 93.36% and a specificity of 93.36%. The feature extraction and classification methods are useful and reliable for sputum detection. This approach differs from waveform research and can provide a better visualization of sputum conditions. The proposed system can be used in the ICU to inform medical staff when sputum is present in a patient's trachea.

A Lung Sound Category Recognition Method Based on Wavelet Decomposition and BP Neural Network

In this paper, a method of characteristic extraction and recognition on lung sounds is given. Wavelet de-noised method is adopted to reduce noise of collected lung sounds and extract wavelet characteristic coefficients of the de-noised lung sounds by wavelet decomposition. Considering the problem that lung sounds characteristic vectors are of high dimensions after wavelet decomposition and reconstruction, a new method is proposed to transform the characteristic vectors from reconstructed signals into reconstructed signal energy. In addition, we use linear discriminant analysis (LDA) to reduce the dimension of characteristic vectors for comparison in order to obtain a more efficient way for recognition. Finally, we use BP neural network to carry out lung sounds recognition where comparatively high-dimensional characteristic vectors and low- dimensional vectors are set as input and lung sounds categories as output with a recognition accuracy of 82.5% and 92.5%.

Evolutionary Algorithm-Based Friction Feedforward Compensation for a Pneumatic Rotary Actuator Servo System

The friction interference in the pneumatic rotary actuator is the primary factor affecting the position accuracy of a pneumatic rotary actuator servo system. The paper proposes an evolutionary algorithm-based friction-forward compensation control architecture for improving position accuracy. Firstly, the basic equations of the valve-controlled actuator are derived and linearized in the middle position, and the transfer function of the system is further obtained. Then, the evolutionary algorithm-based friction feedforward compensation control architecture is structured, including that the evolutionary algorithm is used to optimize the controller coefficients and identify the friction parameters. Finally, the contrast experiments of four control strategies (the traditional PD control, the PD control with friction feedforward compensation without evolutionary algorithm tuning, the PD control with friction feedforward compensation based on the differential evolution algorithm, and the PD control with friction feedforward compensation based on the genetic algorithm) are carried out on the experimental platform. The experimental results reveal that the evolutionary algorithm-based friction feedforward compensation greatly improves the position tracking accuracy and positioning accuracy, and that the differential evolution-based case achieves better accuracy. Also, the system with the friction feedforward compensation still maintains high accuracy and strong stability in the case of load.

Demonstrating the Potential of Using Transcutaneous Oxygen and Carbon Dioxide Tensions to Assess the Risk of Pressure Injuries

Pressure injuries have a high incidence in elderly and critically ill patients, and can endanger lives in severe cases. The key to reducing the incidence of pressure injuries is to find an objective, noninvasive, automatic and consistent scientific method for assessing pressure injuries. To serve this need, we conducted a clinical study to investigate the potential of using transcutaneous oxygen tension (TcPO2) and transcutaneous carbon dioxide tension (TcPCO2) for assessing pressure injuries. From the results of the study we found that first, the values of TcPO2 and TcPCO2 are sensitive to the change of pressure imposed on the measured region and to the risk status of a pressure injury when a pressure is imposed. Second, the magnitude of change in TcPO2 and TcPCO2 is higher in patients with a high risk of a pressure injury compared with those who have a low risk. Third, TcPO2 and TcPCO2 are both significantly correlated with the Braden score, the widely used score for assessing the risk of a pressure injury. Therefore, TcPO2 and TcPCO2 have a potential to be an effective and convenient scientific tool for assessing the risk of pressure injuries.

Entropy Change of Biological Dynamics in Asthmatic Patients and Its Diagnostic Value in Individualized Treatment: A Systematic Review

Asthma is a chronic respiratory disease featured with unpredictable flare-ups, for which continuous lung function monitoring is the key for symptoms control. To find new indices to individually classify severity and predict disease prognosis, continuous physiological data collected from monitoring devices is being studied from different perspectives. Entropy, as an analysis method for quantifying the inner irregularity of data, has been widely applied in physiological signals. However, based on our knowledge, there is no such study to summarize the complexity differences of various physiological signals in asthmatic patients. Therefore, we organized a systematic review to summarize the complexity differences of important signals in patients with asthma. We searched several medical databases and systematically reviewed existing asthma clinical trials in which entropy changes in physiological signals were studied. As a conclusion, we find that, for airflow, heart rate variability, center of pressure and respiratory impedance, their entropy values decrease significantly in asthma patients compared to those of healthy people, while, for respiratory sound and airway resistance, their entropy values increase along with the progression of asthma. Entropy of some signals, such as respiratory inter-breath interval, shows strong potential as novel indices of asthma severity. These results will give valuable guidance for the utilization of entropy in physiological signals. Furthermore, these results should promote the development of management and diagnosis of asthma using continuous monitoring data in the future.

Classification of Sputum Sounds Using Artificial Neural Network and Wavelet Transform

Sputum sounds are biological signals used to evaluate the condition of sputum deposition in a respiratory system. To improve the efficiency of intensive care unit (ICU) staff and achieve timely clearance of secretion in patients with mechanical ventilation, we propose a method consisting of feature extraction of sputum sound signals using the wavelet transform and classification of sputum existence using artificial neural network (ANN). Sputum sound signals were decomposed into the frequency subbands using the wavelet transform. A set of features was extracted from the subbands to represent the distribution of wavelet coefficients. An ANN system, trained using the Back Propagation (BP) algorithm, was implemented to recognize the existence of sputum sounds. The maximum precision rate of automatic recognition in texture of signals was as high as 84.53%. This study can be referred to as the optimization of performance and design in the automatic technology for sputum detection using sputum sound signals.

Pneumatic Rotary Actuator Position Servo System Based on ADE-PD Control

In order to accurately control the rotation position of a pneumatic rotary actuator, the flow state of the gas and the motion state of the pneumatic rotary actuator in the pneumatic rotary actuator position servo system are analyzed in this paper. The mathematical model of the system and the experiment platform are established after that. An Adaptive Differential Evolution (ADE) algorithm which adaptively ameliorates the scaling factor and crossover probability in the process of individual evolution is proposed and applied to the parameter optimization of PD controller. The experimental platform is used to compare the controller with Differential Evolution (DE) algorithm and NCD-PID controller. Finally, the characteristics of the system are tested by increasing the inertial load. The experimental results illustrate that system using ADE-PD control strategy has greater position precision and faster response than using DE-PD and NCD-PID strategies, and shows great robustness.