An adaptive noise reduction stethoscope for auscultation in high noise environments

Coherence Function and Adaptive Noise Cancellation Performance of an Acoustic Sensor System for Use in Detecting Coronary Artery Disease

Adaptive noise cancellation is a useful linear technique to attenuate unwanted background noise that cannot be removed using traditional frequency-selective filters. Usually, this is due to the signal and noise co-existing in the same frequency band. This paper tests a weighted least mean squares (WLMS) algorithm on a stethoscope system for use in detecting coronary artery disease in the presence of background noise. Each stethoscope is equipped with two microphones: one used to detect heart signals and one used to detect background noise. The WLMS method was used for four different sources of background noise whilst measuring a heartbeat, including a single tone, multiple tones, hospital/clinic noise, and breathing noise. The magnitude-squared coherence between both microphones was unity for the tone scenarios, resulting in complete attenuation. For the other background noise sources, a less-than-unity magnitude-squared coherence resulted in minor and no attenuation. Thus, the coherence function is a tool that can be used to predict the amount of attenuation achievable by linear adaptive noise-cancellation techniques, such as WLMS, as presented in this article.

Comprehensive Analysis System for Automated Respiratory Cycle Segmentation and Crackle Peak Detection

Digital auscultation is a well-known method for assessing lung sounds, but remains a subjective process in typical practice, relying on the human interpretation. Several methods have been presented for detecting or analyzing crackles but are limited in their real-world application because few have been integrated into comprehensive systems or validated on non-ideal data. This work details a complete signal analysis methodology for analyzing crackles in challenging recordings. The procedure comprises five sequential processing blocks: (1) motion artifact detection, (2) deep learning denoising network, (3) respiratory cycle segmentation, (4) separation of discontinuous adventitious sounds from vesicular sounds, and (5) crackle peak detection. This system uses a collection of new methods and robustness-focused improvements on previous methods to analyze respiratory cycles and crackles therein. To validate the accuracy, the system is tested on a database of 1000 simulated lung sounds with varying levels of motion artifacts, ambient noise, cycle lengths and crackle intensities, in which ground truths are exactly known. The system performs with average F-score of 91.07% for detecting motion artifacts and 94.43% for respiratory cycle extraction, and an overall F-score of 94.08% for detecting the locations of individual crackles. The process also successfully detects healthy recordings. Preliminary validation is also presented on a small set of 20 patient recordings, for which the system performs comparably. These methods provide quantifiable analysis of respiratory sounds to enable clinicians to distinguish between types of crackles, their timing within the respiratory cycle, and the level of occurrence. Crackles are one of the most common abnormal lung sounds, presenting in multiple cardiorespiratory diseases. These features will contribute to a better understanding of disease severity and progression in an objective, simple and non-invasive way.

Cardiovascular Disease Recognition Based on Heartbeat Segmentation and Selection Process

Assessment of heart sounds which are generated by the beating heart and the resultant blood flow through it provides a valuable tool for cardiovascular disease (CVD) diagnostics. The cardiac auscultation using the classical stethoscope phonological cardiogram is known as the most famous exam method to detect heart anomalies. This exam requires a qualified cardiologist, who relies on the cardiac cycle vibration sound (heart muscle contractions and valves closure) to detect abnormalities in the heart during the pumping action. Phonocardiogram (PCG) signal represents the recording of sounds and murmurs resulting from the heart auscultation, typically with a stethoscope, as a part of medical diagnosis. For the sake of helping physicians in a clinical environment, a range of artificial intelligence methods was proposed to automatically analyze PCG signal to help in the preliminary diagnosis of different heart diseases. The aim of this research paper is providing an accurate CVD recognition model based on unsupervised and supervised machine learning methods relayed on convolutional neural network (CNN). The proposed approach is evaluated on heart sound signals from the well-known, publicly available PASCAL and PhysioNet datasets. Experimental results show that the heart cycle segmentation and segment selection processes have a direct impact on the validation accuracy, sensitivity (TPR), precision (PPV), and specificity (TNR). Based on PASCAL dataset, we obtained encouraging classification results with overall accuracy 0.87, overall precision 0.81, and overall sensitivity 0.83. Concerning Micro classification results, we obtained Micro accuracy 0.91, Micro sensitivity 0.83, Micro precision 0.84, and Micro specificity 0.92. Using PhysioNet dataset, we achieved very good results: 0.97 accuracy, 0.946 sensitivity, 0.944 precision, and 0.946 specificity.

Design and Comparative Performance of a Robust Lung Auscultation System for Noisy Clinical Settings

Chest auscultation is a widely used clinical tool for respiratory disease detection. The stethoscope has undergone a number of transformative enhancements since its invention, including the introduction of electronic systems in the last two decades. Nevertheless, stethoscopes remain riddled with a number of issues that limit their signal quality and diagnostic capability, rendering both traditional and electronic stethoscopes unusable in noisy or non-traditional environments (e.g., emergency rooms, rural clinics, ambulatory vehicles). This work outlines the design and validation of an advanced electronic stethoscope that dramatically reduces external noise contamination through hardware redesign and real-time, dynamic signal processing. The proposed system takes advantage of an acoustic sensor array, an external facing microphone, and on-board processing to perform adaptive noise suppression. The proposed system is objectively compared to six commercially-available acoustic and electronic devices in varying levels of simulated noisy clinical settings and quantified using two metrics that reflect perceptual audibility and statistical similarity, normalized covariance measure (NCM) and magnitude squared coherence (MSC). The analyses highlight the major limitations of current stethoscopes and the significant improvements the proposed system makes in challenging settings by minimizing both distortion of lung sounds and contamination by ambient noise.

Computerized Lung Sound Screening for Pediatric Auscultation in Noisy Field Environments

GOAL

Chest auscultations offer a non-invasive and low-cost tool for monitoring lung disease. However, they present many shortcomings, including inter-listener variability, subjectivity, and vulnerability to noise and distortions. This work proposes a computer-aided approach to process lung signals acquired in the field under adverse noisy conditions, by improving the signal quality and offering automated identification of abnormal auscultations indicative of respiratory pathologies.

METHODS

The developed noise-suppression scheme eliminates ambient sounds, heart sounds, sensor artifacts, and crying contamination. The improved high-quality signal is then mapped onto a rich spectrotemporal feature space before being classified using a trained support-vector machine classifier. Individual signal frame decisions are then combined using an evaluation scheme, providing an overall patient-level decision for unseen patient records.

RESULTS

All methods are evaluated on a large dataset with 1000 children enrolled, 1-59 months old. The noise suppression scheme is shown to significantly improve signal quality, and the classification system achieves an accuracy of 86.7% in distinguishing normal from pathological sounds, far surpassing other state-of-the-art methods.

CONCLUSION

Computerized lung sound processing can benefit from the enforcement of advanced noise suppression. A fairly short processing window size (  s) combined with detailed spectrotemporal features is recommended, in order to capture transient adventitious events without highlighting sharp noise occurrences.

SIGNIFICANCE

Unlike existing methodologies in the literature, the proposed work is not limited in scope or confined to laboratory settings: This work validates a practical method for fully automated chest sound processing applicable to realistic and noisy auscultation settings.

Adaptive Noise Suppression of Pediatric Lung Auscultations With Real Applications to Noisy Clinical Settings in Developing Countries

GOAL

Chest auscultation constitutes a portable low-cost tool widely used for respiratory disease detection. Though it offers a powerful means of pulmonary examination, it remains riddled with a number of issues that limit its diagnostic capability. Particularly, patient agitation (especially in children), background chatter, and other environmental noises often contaminate the auscultation, hence affecting the clarity of the lung sound itself. This paper proposes an automated multiband denoising scheme for improving the quality of auscultation signals against heavy background contaminations.

METHODS

The algorithm works on a simple two-microphone setup, dynamically adapts to the background noise and suppresses contaminations while successfully preserving the lung sound content. The proposed scheme is refined to offset maximal noise suppression against maintaining the integrity of the lung signal, particularly its unknown adventitious components that provide the most informative diagnostic value during lung pathology.

RESULTS

The algorithm is applied to digital recordings obtained in the field in a busy clinic in West Africa and evaluated using objective signal fidelity measures and perceptual listening tests performed by a panel of licensed physicians. A strong preference of the enhanced sounds is revealed.

SIGNIFICANCE

The strengths and benefits of the proposed method lie in the simple automated setup and its adaptive nature, both fundamental conditions for everyday clinical applicability. It can be simply extended to a real-time implementation, and integrated with lung sound acquisition protocols.

In-flight auscultation during medical air evacuation: comparison between traditional and amplified stethoscopes

OBJECTIVES

The aim of this study was to evaluate the capacity of a traditional stethoscope versus an electronically amplified one (expected to reduce background and ambient noise) to assess heart and respiratory sounds during medical transport.

MATERIALS AND METHODS

It was a prospective, double-blinded, randomized performed study. One traditional stethoscope (Littmann Cardiology III; 3M, St Paul, MN) and 1 electronically amplified stethoscope (Littmann 3200, 3M) were used for our tests. Heart and lung auscultation during real medical evacuations aboard a medically configured Falcon 50 aircrafts were studied. The quality of auscultation was ranged using a numeric rating scale from 0 to 10 (0 corresponding to "I hear nothing" and 10 to "I hear perfectly"). Data collected were compared using a t-test for paired values.

RESULTS

A total of 40 comparative evaluations were performed. For cardiac auscultation, the value of the rating scale was 4.53 ± 1.91 and 7.18 ± 1.88 for the traditional and amplified stethoscope, respectively (paired t-test: P < .0001). For respiratory sounds, quality of auscultation was estimated at 3.1 ± 1.95 for a traditional stethoscope and 5.10 ± 2.13 for the amplified one (paired t-test: P < .0001).

CONCLUSIONS

This study showed that practitioners would be better helped in hearing cardiac and respiratory sounds with an electronically amplified stethoscope than with a traditional one during air medical transport in a medically configured Falcon 50 aircraft.

Using the Entropy of Tracheal Sounds to Detect Apnea during Sedation in Healthy Nonobese Volunteers

Background:

Undetected apnea can lead to severe hypoxia, bradycardia, and cardiac arrest. Tracheal sounds entropy has been proved to be a robust method for estimating respiratory flow, thus maybe a more reliable way to detect obstructive and central apnea during sedation.

Methods:

A secondary analysis of a previous pharmacodynamics study was conducted. Twenty volunteers received propofol and remifentinal until they became unresponsive to the insertion of a bougie into the esophagus. Respiratory flow rate and tracheal sounds were recorded using a pneumotachometer and a microphone. The logarithm of the tracheal sound Shannon entropy (Log-E) was calculated to estimate flow rate. An adaptive Log-E threshold was used to distinguish between the presence of normal breath and apnea. Apnea detected from tracheal sounds was compared to the apnea detected from respiratory flow rate.

Results:

The volunteers stopped breathing for 15 s or longer (apnea) 322 times during the 12.9-h study. Apnea was correctly detected 310 times from both the tracheal sounds and the respiratory flow. Periods of apnea were not detected by the tracheal sounds 12 times. The absence of tracheal sounds was falsely detected as apnea 89 times. Normal breathing was detected correctly 1,196 times. The acoustic method detected obstructive and central apnea in sedated volunteers with 95% sensitivity and 92% specificity.

Conclusions:

We found that the entropy of the acoustic signal from a microphone placed over the trachea may reliably provide an early warning of the onset of obstructive and central apnea in volunteers under sedation.

Acoustical analysis of mechanical heart valve sounds for early detection of malfunction

Mechanical heart valves carry the disadvantage of lifelong antithrombotic therapy, due to the high risk of thrombus formation on the valve surface. Current diagnostic methods are incapable of detecting thrombus formation in an early stage. This article investigates a new diagnostic method, based on the analysis of the acoustic signal produced by the valve. This method should be capable of early detection of malfunction, thus permitting targeted medication and reducing valve-related complications and mortality. A measurement setup assuring optimal signal quality was developed, and a signal analysis program was implemented and validated on an in vitro mock circulatory loop. Next, four sheep were implanted with a bileaflet mechanical valve. The signals of their valves developing thrombosis were assessed on a weekly basis before explantation. Three sheep were sacrificed shortly after detection of malfunction according to the newly developed method. In each case, thrombus or membrane formation was detected on the leaflets upon explantation. In one sheep, no malfunction was found in the analysis, which was also confirmed by the condition of the valve upon explantation. These preliminary results indicate that acoustical analysis of mechanical heart valves permits early detection of valvular malfunction. Further research with more in vitro and animal testing is required to statistically validate these findings.

Breath sound changes associated with malpositioned endotracheal tubes

Endotracheal tubes (ETTs) are used to establish airway access in patients with ventilatory failure and during general anaesthesia. Tube malpositioning can compromise respiratory function and can be associated with increased morbidity and mortality. Clinical assessment of ETT position normally involves chest auscultation, which is highly skill-dependent and can be misleading. The objective of this pilot study was to investigate breath sound changes associated with ETT malpositioning. Breath sounds were acquired in six human subjects over each hemithorax and over the epigastrium for tracheal, bronchial and oesophageal intubations. When the ETT was in the oesophagus, the acoustic energy ratio between epigastrium and chest surface increased in all subjects (p < 0.04). In addition, ETT placement in the right mainstem bronchus decreased the acoustic energy ratio between the left and right hemithoraxes in all subjects (p < 0.04). A baseline measurement of this energy ratio was needed for bronchial intubation identification. However, using this ratio after bandpass filtering (200-500 Hz) did not require a baseline value, which would increase the utility of this method for initial ETT placement. These results suggest that computerised analysis of breath sounds may be useful for assessment of ETT positioning. More studies are needed to test the feasibility of this approach further.

The provision of sophisticated critical care beyond the hospital: Lessons from physiology and military experiences that apply to civil disaster medical response

OBJECTIVE

The provision of sophisticated medical care in an austere environment is challenging. During and after a mass casualty event, it is likely that critical care services will be needed beyond an intensive care unit (ICU) setting. The objective of this article is to explore existing ICU care systems such as military aeromedical transport that may be applicable to disaster medicine and to providing critical care outside of an ICU setting.

RESULTS

The U.S. Air Force Critical Care Aeromedical Transport (CCAT) Teams were developed in 1994 in response to an unmet military need for long-range air transport of critically ill and injured patients. This system has transported several thousand ICU patients and is an applicable model for the future development of extrahospital critical care capabilities needed during a disaster. We also discuss civilian aeromedical critical care systems, the types of medical devices used, and their applicability to disaster medical response.

CONCLUSION

The U.S. Air Force CCAT Team program, as well as many civilian critical care air ambulance services, provides a workable starting point for the development of disaster medical critical care response capabilities for disaster medical systems.

Representation and classification of breath sounds recorded in an intensive care setting using neural networks

OBJECTIVE

Develop and test methods for representing and classifying breath sounds in an intensive care setting.

METHODS

Breath sounds were recorded over the bronchial regions of the chest. The breath sounds were represented by their averaged power spectral density, summed into feature vectors across the frequency spectrum from 0 to 800 Hertz. The sounds were segmented by individual breath and each breath was divided into inspiratory and expiratory segments. Sounds were classified as normal or abnormal. Different back-propagation neural network configurations were evaluated. The number of input features, hidden units, and hidden layers were varied.

RESULTS

2127 individual breath sounds from the ICU patients and 321 breaths from training tapes were obtained. Best overall classification rate for the ICU breath sounds was 73% with 62% sensitivity and 85% specificity. Best overall classification rate for the training tapes was 91% with 87% sensitivity and 95% specificity.

CONCLUSIONS

Long term monitoring of lung sounds is not feasible unless several barriers can be overcome. Several choices in signal representation and neural network design greatly improved the classification rates of breath sounds. The analysis of transmitted sounds from the trachea to the lung is suggested as an area for future study.

Use of abdominal percussion for pneumoperitoneum detection

Pneumoperitoneum refers to free air within the abdominal cavity that typically signifies serious abdominal pathology such as a perforated gut. The principal hypothesis of the study was that abdominal structure alterations due to pneumoperitoneum cause diagnostic changes in the sounds induced by abdominal percussion. The current pilot study investigated these changes in a mongrel dog model. Abdominal percussion was performed at baseline and after creation of pneumoperitoneum states. The resulting acoustic events were acquired, digitised and analysed. The event attack and decay rates and dominant frequencies during decay decreased with pneumoperitoneum (p = 0.084, 0.014 and 0.004, respectively; Wilcoxon signed-rank test). Simple theoretical models were constructed and predicted the observed decrease in resonant frequencies with increasing air pocket size. The results suggested that the normal and the 1,000 ml pneumoperitoneum states can be separated using thresholds of the attack and decay rates and resonant frequency (specificity = 80%, 100% and 100%, and sensitivity = 100%, 100% and 100%, respectively). Separating the control and the 500 ml pneumoperitoneum cases may be also possible (specificity = 80%, 100%, 100% and sensitivity = 50%, 70% and 90%, respectively), but separating the two levels of pneumoperitoneum was not feasible using the current approach. Therefore analysis of abdominal percussion sounds may prove useful for pneumoperitoneum detection, but not for distinguishing different levels of that condition.

Evaluation of a noise reduction system for the assessment of click-evoked otoacoustic emissions

A recently proposed noise reduction system intended to facilitate the assessment of click-evoked otoacoustic emission (CEOAE) in noisy environments [Comput. Biol. Med. 30, 341 (2000)] is evaluated using 13 normally hearing ears and 9 ears with a sensorineural hearing loss. The noise reduction system is based on an adaptive noise canceller design using an additional noise-only reference microphone and intended to reduce externally generated noise. The system is tested in quiet and at different levels of white noise. The three main design parameters of the noise reduction system (adaptation time constant, length of the adaptive filter, and position of the noise reference microphone) are varied systematically in different experiments. With the noise reduction system active, CEOAE can be assessed correctly at noise levels which are 5 to 9 dB higher than without the noise reduction system. For the range of adaptation time constants considered (65.6 to 656 ms), no statistically significant effect on the amount of noise reduction is observed. Noise reduction is highest when the reference microphone is positioned close to the ear probe. Using this reference microphone position and adaptive filters of 6.56 ms in length, average noise reductions of 7.17 to 8.50 dB are achieved.