Abnormal Heart Sounds Research Articles

Improved Segmentation with Dynamic Threshold Adjustment for Phonocardiography Recordings.

In this study with the intent to segment phonocardiography (PCG) recordings, an algorithm that processes a feature vector obtained by jointly using wavelet transform and mel scaled energy spectrum of the PCG signal is proposed. The feature vector is processed by a peak detection algorithm which results in a set of peaks that meet some certain criteria and will be exploited in the succeeding stages of the algorithm. Heart sounds are labeled by convolving circularly a template and a fragment of the feature vector which is picked up with the guidance of the peaks. At the final stage, the algorithm tries to detect and correct erroneous labels. The performance of the algorithm was tested on both normal heart sounds and abnormal heart sounds; 80 records in total. As a result of these tests, for normal heart sounds S1 and S2 sounds were detected with 99,51% recall and 97,28% precision, while with 97,59% recall and 92,53% precision for abnormal heart sounds.

Cepstral Features Extraction for Heart Sounds Classification

A novel method for separation between normal and abnormal heart sounds based on Phonocardiogram (PCG) is presented in this article. For features extraction phase, Mel-Frequency Cepstral Coefficients (MFCC) algorithm is used to extract information from heart sound signals. In this step, changing the frames size, during framing process, shows it influence on the obtained result. In classification step, Support Vector Machines (SVM) is used with different kernels. Simulation results obtained from different databases are compared and discussed. The developed system gave good results when applied to different datasets with an accuracy of 96% and 98% for the first and the second dataset respectively. The developed system can be used for other applications in biomedical domains

A qualitative exploration of current paramedic cardiac auscultation practices

This exploratory study addresses the current paucity of knowledge available in UK paramedic practice in relation to cardiac auscultation. There is a recognised lack of data surrounding the efficacy, safety and relevance of patient assessment skills in the pre-hospital setting in general, and cardiac auscultation specifically. This study provides information about current paramedic practice, and provides a basis for further research in this area. An online survey was distributed using convenience and snowball sampling, receiving 328 responses within a 31-day period. The results show that many paramedics rarely, or indeed never, undertake cardiac auscultation and that many lack confidence in recognising normal and abnormal heart sounds. There is also a divided opinion among respondents who provided free-text answers, with some feeling that the skill of cardiac auscultation is vital in pre-hospital care and others firmly disagreeing. This research lays the groundwork for further developments in training, education and continuing professional development for paramedics.

Analysis of PCG signals using quality assessment and homomorphic filters for localization and classification of heart sounds

Background and Objective: Accurate localization of heart beats in phonocardiogram (PCG) signal is very crucial for correct segmentation and classification of heart sounds into S1 and S2. This task becomes challenging due to inclusion of noise in acquisition process owing to number of different factors. In this paper we propose a system for heart sound localization and classification into S1 and S2. The proposed system introduces the concept of quality assessment before localization, feature extraction and classification of heart sounds. Methods: The signal quality is assessed by predefined criteria based upon number of peaks and zero crossing of PCG signal. Once quality assessment is performed, then heart beats within PCG signal are localized, which is done by envelope extraction using homomorphic envelogram and finding prominent peaks. In order to classify localized peaks into S1 and S2, temporal and time-frequency based statistical features have been used. Support Vector Machine using radial basis function kernel is used for classification of heart beats into S1 and S2 based upon extracted features. The performance of the proposed system is evaluated using Accuracy, Sensitivity, Specificity, F-measure and Total Error. The dataset provided by PASCAL classifying heart sound challenge is used for testing. Results: Performance of system is significantly improved by quality assessment. Results shows that proposed Localization algorithm achieves accuracy up to 97% and generates smallest total average error among top 3 challenge participants. The classification algorithm achieves accuracy up to 91%. Conclusion: The system provides firm foundation for the detection of normal and abnormal heart sounds for cardiovascular disease detection.

Deep Unsupervised Representation Learning for Abnormal Heart Sound Classification.

Given the world-wide prevalence of heart disease, the robust and automatic detection of abnormal heart sounds could have profound effects on patient care and outcomes. In this regard, a comparison of conventional and state-of-theart deep learning based computer audition paradigms for the audio classification task of normal, mild abnormalities, and moderate/severe abnormalities as present in phonocardiogram recordings, is presented herein. In particular, we explore the suitability of deep feature representations as learnt by sequence to sequence autoencoders based on the auDeep toolkit. Key results, gained on the new Heart Sounds Shenzhen corpus, indicate that a fused combination of deep unsupervised features is well suited to the three-way classification problem, achieving our highest unweighted average recall of 47.9% on the test partition.

Spectral Fault Recovery Analysis Revisited With Normal and Abnormal Heart Sound Signals

The computation of fast Fourier transform (FFT) plays a vital role in all fields of science and engineering, especially in the medical domain. Spectral results of the data convey more information to the clinicians to take handy decisions during the diagnosis. If a problem happens during the computation of FFT, the incorrect results misinterpret the information and hence the suggestions and decisions by practitioners would be the useless one moreover, it may yield harmful result to the patients. This paper deals with how to recover the spectrum from the faulty one with the heart sound case reports. The heart sound recording is the phonocardiogram (PCG) signal. Stethoscopes are commonly used to hear the heart sound for diagnosis. The two PCG signals of normal and abnormal subjects in each were analyzed in this paper. The normal subject produce clear lub and dub sounds, where the abnormal subject produce a kind of whistling or swishing sound in middle which may be due to the problem of septal defect in a heart. This paper analyzes four PCG signals spectrally with fault and the recovered ones along with the execution time results in each using Matlab R2016b tool.

How to maximise your auscultation technique

Auscultation is a cheap and easy diagnostic tool available in veterinary practice. Yet it often creates uncertainty, prompting both nurses and veterinary surgeons to seek second opinions amongst colleagues. The purpose of this article is to explore the best techniques for auscultation, and discuss the different sounds that can be heard in dogs and cats when listening to the heart. The starting point of auscultation is to identify normal heart sounds. This will then help recognise abnormal heart sounds, which can then be split into loudness, timing, and point of maximal intensity.

Automatic heart activity diagnosis based on Gram polynomials and probabilistic neural networks.

The paper proposes a new approach to heart activity diagnosis based on Gram polynomials and probabilistic neural networks (PNN). Heart disease recognition is based on the analysis of phonocardiogram (PCG) digital sequences. The PNN provides a powerful tool for proper classification of the input data set. The novelty of the proposed approach lies in a powerful feature extraction based on Gram polynomials and the Fourier transform. The proposed system presents good performance obtaining overall sensitivity of 93%, specificity of 91% and accuracy of 94%, using a public database of over 3000 heart beat sound recordings, classified as normal and abnormal heart sounds. Thus, it can be concluded that Gram polynomials and PNN prove to be a very efficient technique using the PCG signal for characterizing heart diseases.

A Case of Methemoglobinemia Successfully Treated with Hyperbaric Oxygenation Monotherapy

BackgroundMethylene blue is the first-line therapy for methemoglobinemia, but it can be intermittently unavailable due to production issues. For this clinical scenario, alternative treatment options need to be explored. Hyperbaric oxygenation (HBO) is conventionally applied as an adjunctive therapy during the systemic administration of methylene blue. Currently, little is known regarding the effects of HBO monotherapy in methemoglobinemia. We report a case of methemoglobinemia that was successfully treated with HBO monotherapy. Case ReportA 41-year-old man presented to the Emergency Department with dyspnea and dizziness subsequent to smoking in a garage filled with motor vehicle exhaust gas. There were no abnormal heart or lung sounds. While administering oxygen flowing at 15 L/min via a mask with a reservoir bag, blood tests revealed high methemoglobin (MetHb) levels at 59.6%. He was treated with HBO monotherapy, and sequential tests showed that the MetHb level decreased significantly to 34.0%, 12.8%, 6.2%, and eventually, 3.5%. He was discharged with stable vital signs the next day. Why Should an Emergency Physician Be Aware of This?HBO monotherapy is an effective alternative treatment for methemoglobinemia when methylene blue is not available.

DropConnected neural networks trained on time-frequency and inter-beat features for classifying heart sounds

Objective: Automatic heart sound analysis has the potential to improve the diagnosis of valvular heart diseases in the primary care phase, as well as in countries where there is neither the expertise nor the equipment to perform echocardiograms. An algorithm has been trained, on the PhysioNet open-access heart sounds database, to classify heart sounds as normal or abnormal. Approach: The heart sounds are segmented using an open-source algorithm based on a hidden semi-Markov model. Following this, the time-frequency behaviour of a single heartbeat is characterized by using a novel implementation of the continuous wavelet transform, mel-frequency cepstral coefficients, and certain complexity measures. These features help detect the presence of any murmurs. A number of other features are also extracted to characterise the inter-beat behaviour of the heart sounds, which helps to recognize diseases such as arrhythmia. The extracted features are normalized and their dimensionality is reduced using principal component analysis. They are then used as the input to a fully-connected, two-hidden-layer neural network, trained by error backpropagation, and regularized with DropConnect.Main results: This algorithm achieved an accuracy of 85.2% on the test data, which placed third in the PhysioNet/Computing in Cardiology Challenge (first place scored 86.0%). However, this is unrealistic of real-world performance, as the test data contained a dataset (dataset-e) in which normal and abnormal heart sounds were recorded with different stethoscopes. A 10-fold cross-validation study on the training data (excluding dataset-e) gives a mean score of 74.8%, which is a more realistic estimate of accuracy. With dataset-e excluded from training, the algorithm scored only 58.1% on the test data.

Heart sounds analysis using probability assessment

Objective: This paper describes a method for automated discrimination of heart sounds recordings according to the Physionet Challenge 2016. The goal was to decide if the recording refers to normal or abnormal heart sounds or if it is not possible to decide (i.e. ‘unsure’ recordings). Approach: Heart sounds S1 and S2 are detected using amplitude envelopes in the band 15–90 Hz. The averaged shape of the S1/S2 pair is computed from amplitude envelopes in five different bands (15–90 Hz; 55–150 Hz; 100–250 Hz; 200–450 Hz; 400–800 Hz). A total of 53 features are extracted from the data. The largest group of features is extracted from the statistical properties of the averaged shapes; other features are extracted from the symmetry of averaged shapes, and the last group of features is independent of S1 and S2 detection. Generated features are processed using logical rules and probability assessment, a prototype of a new machine-learning method. Main results: The method was trained using 3155 records and tested on 1277 hidden records. It resulted in a training score of 0.903 (sensitivity 0.869, specificity 0.937) and a testing score of 0.841 (sensitivity 0.770, specificity 0.913). The revised method led to a test score of 0.853 in the follow-up phase of the challenge. Significance: The presented solution achieved 7th place out of 48 competing entries in the Physionet Challenge 2016 (official phase). In addition, the PROBAfind software for probability assessment was introduced.

A robust dataset-agnostic heart disease classifier from Phonocardiogram.

Automatic classification of normal and abnormal heart sounds is a popular area of research. However, building a robust algorithm unaffected by signal quality and patient demography is a challenge. In this paper we have analysed a wide list of Phonocardiogram (PCG) features in time and frequency domain along with morphological and statistical features to construct a robust and discriminative feature set for dataset-agnostic classification of normal and cardiac patients. The large and open access database, made available in Physionet 2016 challenge was used for feature selection, internal validation and creation of training models. A second dataset of 41 PCG segments, collected using our in-house smart phone based digital stethoscope from an Indian hospital was used for performance evaluation. Our proposed methodology yielded sensitivity and specificity scores of 0.76 and 0.75 respectively on the test dataset in classifying cardiovascular diseases. The methodology also outperformed three popular prior art approaches, when applied on the same dataset.

Auscultation of the heart: The Basics with Anatomical Correlation.

Proficiency in heart auscultation continues to be important even in an era of modern technology. However, many physicians and health care providers are uncomfortable and often inaccurate in their ability to identify normal and abnormal heart sounds. The following review was performed to provide a basis that health care providers can use to strengthen their understanding and improve their techniques when auscultating the human heart. Clin. Anat. 30:58-60, 2017. © 2016 Wiley Periodicals, Inc.

3D heart sound source localization via combinational subspace methods for long-term heart monitoring

Abstract Heart sound is a key mean for heart monitoring and a common known early diagnosis modality. In this paper, by employing a multiple channel sound recording system and implementing a spatiotemporal information extraction process, a powerful approach is proposed for heart monitoring. Use of the extracted spatiotemporal information of the heart sound, which in turn could lead to detection of the time duration and the heart abnormalities source locations, gives a clinical contribution to the diagnosis of heart valvular diseases. Since the heart valves have very small distances from each other and are located in a reverberant environment of chest and its surrounding muscles, we take the advantages of the group-delay information of the Recursively Applied and Projected-MUltiple SIgnal Classification (RAP-MUSIC) method to localize the active heart valves in the S1, systole, S2, and diastole phases of the heart sound. The proposed algorithm is applied to a simulated database and compared with the MUltiple SIgnal Classification (MUSIC), RAPMUSIC and group delayed MUSIC algorithms. The comparison shows the proposed combinational method has the highest resolution in the simulated environment and more accuracy when sources are very close. In the following, the proposed algorithm is applied to some normal and abnormal heart sounds recorded by the employed microphone array. The obtained results indicate that the proposed method can localize the heart valves in real conditions. Moreover, it is shown that via this approach, the extra abnormal sound sources in the heart of the patients, if exist, could be properly detected and localized. This 3D heart valve localization method is an appropriate approach suitable to be employed in long-term heart monitoring instruments.

A sneak peek into digital innovations and wearable sensors for cardiac monitoring.

Many mobile phone or tablet applications have been designed to control cardiovascular risk factors (obesity, smoking, sedentary lifestyle, diabetes and hypertension) or to optimize treatment adherence. Some have been shown to be useful but the long-term benefits remain to be demonstrated. Digital stethoscopes make easier the interpretation of abnormal heart sounds, and the development of pocket-sized echo machines may quickly and significantly expand the use of ultrasounds. Daily home monitoring of pulmonary artery pressures with wireless implantable sensors has been shown to be associated with a significant decrease in hospital readmissions for heart failure. There are more and more non-invasive, wireless, and wearable sensors designed to monitor heart rate, heart rate variability, respiratory rate, arterial oxygen saturation, and thoracic fluid content. They have the potential to change the way we monitor and treat patients with cardiovascular diseases in the hospital and beyond. Some may have the ability to improve quality of care, decrease the number of medical visits and hospitalization, and ultimately health care costs. Validation and outcome studies are needed to clarify, among the growing number of digital innovations and wearable sensors, which tools have real clinical value.

Active CMV and EBV infections in renal transplant recipients with unexplained fever and elevated serum creatinine

Proper identification of active cytomegalovirus (CMV) and Epstein-Barr virus (EBV) infections are helpful for monitoring antiviral treatment in transplant recipients. Qualitative and quantitative CMV, EBV DNA PCR techniques in the context of serological tests are performed for early detection and differentiation of active and latent CMV and EBV infections in renal transplantation. Basically, 129 renal transplanted recipients monitored carefully and hospitalized for unexplained elevated creatinine levels or high fever and 21 of their donors were studied. CMV DNA was detected in 63.5% of the febrile episodes following transplantation and in 46.42% of readmitted patients using qualitative PCR method. In the first group, 15% of the patients and in the second group 42.85% of the patients had copy numbers more than cutoff point (900 copies/mL). Cutoff point had 100% sensitivity and 82.5% specificity for active and symptomatic CMV infection. Only 15.5% of the subjects were positive for EBV infection by qualitative PCR method. Among them 5% had >2000 copies/mL and were symptomatic. One subject with a history of three times hospitalization had higher EBV viral load and developed post-transplant lymphoproliferative disorder. CMV load was significantly correlated with elevated creatinine levels (OR = 3.1, p = 0.006), abnormal heart sounds (OR = 4.7; p = 0.02) and hypertension (OR = 3.6; p = 0.03). Only qRT-PCR could differentiate between latent and active infections and might be clinically useful for monitoring symptomatic CMV and EBV infections and initiation of the antiviral therapy. Elevated creatinine levels, hypertension, and abnormal heart sounds could be considered as main manifestations of HCMV infection in kidney recipients.

Use of Harvey® the Cardiopulmonary Patient Simulator in Physician Assistant Training.

The purpose of this study was to evaluate physician assistant students' confidence levels in detection of heart murmurs following instruction with Harvey(R) the Cardiopulmonary Patient Simulator compared with a classroom heart sounds activity. Cohort 1 (n = 33) participated in the classroom heart sounds activity and then participated in the Harvey simulation exercise. Cohort 2 (n = 34) first participated in the Harvey simulation activity and then in the classroom heart sounds activity. All students completed preintervention and postintervention surveys to assess confidence in detecting heart sounds. A multiple-choice quiz was distributed to each group after participation in the first heart sounds activity. Sixty-seven students completed all surveys. Before either activity, 6% of students in Cohort 1 and 3% in Cohort 2 reported confidence in detecting abnormal heart sounds. After completing the first activity, 85% of the classroom heart sounds activity group (Cohort 1) and 53% of the Harvey simulation group (Cohort 2) reported confidence in detecting abnormal heart sounds. The mean score on the multiple-choice quiz was 62% in Cohort 1 and 24% in Cohort 2. Both cohorts reported confidence in learning abnormal heart sounds after participation in the Harvey simulation compared with baseline confidence. Students who participated in the classroom heart sounds activity before the Harvey simulation activity performed higher on the murmur identification multiple-choice quiz. The University of Texas Medical Branch PA faculty should consider continued use of both the classroom heart sounds activity and Harvey simulation.

Analysis of heart sounds using time-frequency visual representations

Phonocardiogram (PCG) is a pre-eminent tool to visualise the normal and abnormal heart sounds which helps the physician in the early detection of cardiac murmurs. PCG is a non-stationary signal with the rapid changes in frequency within one cardiac cycle. Hence, the choice of proper segmentation technique is a crucial factor in determining the success or failure of expert systems in the automated analysis of PCG signals. The normal and abnormal PCG signals from the publicly available database were used in this study. Fast Fourier Transform (FFT) and Discrete Wavelet Transform (DWT) were implemented to visualise the various components in the PCG signals. When FFT was applied to the normal and abnormal PCG signals, the resultant spectrums for both the signals were not much distinguishable. S1 and S2 heart sounds had overlapping frequencies. Moreover, preserving the time and frequency resolution is imperative in analysing the abnormal signals. DWT was implemented and visualised using time-frequency indicators. The study concludes that the DWT is more preferred for analysing the non-stationary PCG signal and it would be a less expensive technique to visualise the heart sounds. Simultaneous recording of ECG and PCG signal will provide a better visualisation of cardiac murmurs.

Detection and identification of S1 and S2 heart sounds using wavelet decomposition method

Phonocardiogram (PCG), the digital recording of heart sounds is becoming increasingly popular as a primary detection system for diagnosing heart disorders and it is relatively inexpensive. Electrocardiogram (ECG) is used during the PCG in order to identify the systolic and diastolic parts manually. In this study a heart sound segmentation algorithm has been developed which separates the heart sound signal into these parts automatically. This study was carried out on 100 patients with normal and abnormal heart sounds. The algorithm uses discrete wavelet decomposition and reconstruction to produce PCG intensity envelopes and separates that into four parts: the first heart sound, the systolic period, the second heart sound and the diastolic period. The performance of the algorithm has been evaluated using 14,000 cardiac periods from 100 digital PCG recordings, including normal and abnormal heart sounds. In tests, the algorithm was over 93% correct in detecting the first and second heart sounds. The presented automatic segmentation algorithm using wavelet decomposition and reconstruction to select suitable frequency band for envelope calculations has been found to be effective to segment PCG signals into four parts without using an ECG.

A Case of Esophageal Thermal Injury Following Radiofrequency Catheter Ablation Therapy

Introduction: Radiofrequency ablation (RFA) is a catheter based ablation technique used in patients with atrial fibrillation(AF) to achieve rhythm control.There is risk of esophageal thermal injury during RFA to the posterior left atrium given its close proximity to the anterior esophagus.We describe a case of thermal injury to the esophagus following RFA therapy to the left atrium. Case Report: A 69-year-old female with a history of hypertension, AF was seen with a two day history of melena.The patient's melenic stools began one day following RFA therapy.The patient reported a “gassy feeling” in her stomach but denied abdominal pain,nausea,vomiting,chest pain,heartburn,dysphagia,o dynophagia,fever/chills.No history of GI bleed or change in stool consistency/color were reported in the past.On exam patient was afebrile,hypertensive(178/78) with heart rate of 80. Abdomen was soft/nontender/nondistended with normoactive bowel sounds, no abdominal masses, organomegaly. Lungs were clear to ausculation as no abnormal heart sounds were noted. Lab results demonstrated Potassium 3.5mmol/L,Sodium 142mmol/L,Chloride 102mmol/L,hemoglobin 13.9 g/dL,platelet count 219 10(3)/mcl,PT 15.6, INR 1.4,PTT 31.5.Liver enzymes were normal.An EGD demonstrated a 6-7 mm healing ulcer in the mid esophagus with no evidence of active bleeding.The submucosa was not seen and no evidence of perforation was noted.She was discharged home with pantoprazole 40mg twice daily,sucralfate 1g four times daily and instructions to follow up in the GI clinic. Discusion: Esophageal injury is one complication of RFA therapy.This is due to the close anatomical proximity between the esophagus and the posterior LA wall which is estimated to be ≤ 5 mm in 40% of the population.The intestinal tissue is also far more susceptible to radiofrequency induced thermal injury than the muscular tissue of the heart. Therefore,direct contact of a probe is not required for thermal injury, as convection heat generated within the LA is enough energy to cause esophageal injury.Early diagnosis is vital to prevent fatal outcomes from complication such as perforation and fistula formation. Atrial fibrillation is projected to surge 2.5-fold in the near future due to the aging demographic leading to an increase use of RFA therapy. This may result in a higher incident of RFA related thermal esophageal injury.Further research is needed to identify risk factors,preventive strategies and optimal treatment for esophageal thermal injury related to left atrial RFA.Figure 1