DigiScope--unobtrusive collection and annotating of auscultations in real hospital environments

The CirCor DigiScope Dataset: From Murmur Detection to Murmur Classification

Cardiac auscultation is one of the most cost-effective techniques used to detect and identify many heart conditions. Computer-assisted decision systems based on auscultation can support physicians in their decisions. Unfortunately, the application of such systems in clinical trials is still minimal since most of them only aim to detect the presence of extra or abnormal waves in the phonocardiogram signal, i.e., only a binary ground truth variable (normal vs abnormal) is provided. This is mainly due to the lack of large publicly available datasets, where a more detailed description of such abnormal waves (e.g., cardiac murmurs) exists. To pave the way to more effective research on healthcare recommendation systems based on auscultation, our team has prepared the currently largest pediatric heart sound dataset. A total of 5282 recordings have been collected from the four main auscultation locations of 1568 patients, in the process, 215780 heart sounds have been manually annotated. Furthermore, and for the first time, each cardiac murmur has been manually annotated by an expert annotator according to its timing, shape, pitch, grading, and quality. In addition, the auscultation locations where the murmur is present were identified as well as the auscultation location where the murmur is detected more intensively. Such detailed description for a relatively large number of heart sounds may pave the way for new machine learning algorithms with a real-world application for the detection and analysis of murmur waves for diagnostic purposes.

Machine Learning and IoT Applied to Cardiovascular Diseases Identification through Heart Sounds: A Literature Review

This article presents a systematic mapping study dedicated to conduct a literature review on machine learning and IoT applied in the identification of diseases through heart sounds. This research was conducted between January 2010 and July 2021, considering IEEE Xplore, PubMed Central, ACM Digital Library, JMIR—Journal of Medical Internet Research, Springer Library, and Science Direct. The initial search resulted in 4372 papers, and after applying the inclusion and exclusion criteria, 58 papers were selected for full reading to answer the research questions. The main results are: of the 58 articles selected, 46 (79.31%) mention heart rate observation methods with wearable sensors and digital stethoscopes, and 34 (58.62%) mention care with machine learning algorithms. The analysis of the studies based on the bibliometric network generated by the VOSviewer showed in 13 studies (22.41%) a trend related to the use of intelligent services in the prediction of diagnoses related to cardiovascular disorders.

A comparative study of electronic stethoscopes for cardiac auscultation

There are several electronic stethoscopes available on the market today, with a very high potential for healthcare namely telemedicine, assisted decision and education. However, there are no recent comparatives studies published about the recording quality of auscultation sounds. In this study we aim to: a) define a ranking, according to experts opinion of 6 of the most relevant electronic stethoscopes on the market today; b) verify if there are any relations between a stethoscope's performance and the type of pathology present; c) analyze if some pathologies are more easily identified than others when using electronic auscultation. Our methodology consisted in creating two study groups: the first group included 18 cardiologists and cardiology house officers, acting as the gold standard of this work. The second included 30 medical students. Using a database of heart sounds recorded in real hospital environments, we applied questionnaires to observers from each group. The first group listened to 60 cardiac auscultations recorded by the 6 stethoscopes, and each one was asked to identify the pathological sound present: aortic stenosis, mitral regurgitation or normal. The second group was asked to choose, between two auscultation recordings, using as criteria the best sound quality for the identification of pathological sounds. Results include a total of 1080 evaluations, in which 72% of cases were correctly diagnosed. A detailed breakdown of these results is presented in this paper. As conclusions, results showed that the impact of the differences between stethoscopes is very small, given that we did not find statistically significant differences between all pairs of stethoscopes. Normal sounds showed to be easier to identify than pathological sounds, but we did not find differences between stethoscopes in this identification.

Adaptive Sojourn Time HSMM for Heart Sound Segmentation

Heart sounds are difficult to interpret due to events with very short temporal onset between them (tens of milliseconds) and dominant frequencies that are out of the human audible spectrum. Computer-assisted decision systems may help but they require robust signal processing algorithms. In this paper, we propose a new algorithm for heart sound segmentation using a hidden semi-Markov model. The proposed algorithm infers more suitable sojourn time parameters than those currently suggested by the state of the art, through a maximum likelihood approach. We test our approach over three different datasets, including the publicly available PhysioNet and Pascal datasets. We also release a pediatric dataset composed of 29 heart sounds. In contrast with any other dataset available online, the annotations of the heart sounds in the released dataset contain information about the beginning and the ending of each heart sound event. Annotations were made by two cardiopulmonologists. The proposed algorithm is compared with the current state of the art. The results show a significant increase in segmentation performance, regardless the dataset or the methodology presented. For example, when using the PhysioNet dataset to train and to evaluate the HSMMs, our algorithm achieved average an F-score of [Formula: see text] compared to [Formula: see text] achieved by the algorithm described in [D.B. Springer, L. Tarassenko, and G. D. Clifford, "Logistic regressionHSMM-based heart sound segmentation," IEEE Transactions on Biomedical Engineering, vol. 63, no. 4, pp. 822-832, 2016]. In this sense, the proposed approach to adapt sojourn time parameters represents an effective solution for heart sound segmentation problems, even when the training data does not perfectly express the variability of the testing data.

The effect of data exchange protocols on decision support systems for heart sounds

Heart auscultation is one of the basic exams performed during a patient physical examination, but it is also one that has a high skill ceiling. Decision support systems can provide physicians with a tool that can help to reduce the demanding skill requirements of this exam. Nevertheless, this second opinion needs to be delivered in a timely interval in order to be truly useful for a physician. To do this we need not only optimized algorithms, but also a well designed system. In this paper, we have studied how two different data exchange protocols, that define how data should be transferred from an acquisition to a process module, can impact the celerity of delivering second opinion to a physician. With data collected from real exams, acquired in a field hospital initiative in Brazil, we recreated two use cases that allowed us to measure performance in the form of time and resources spent, as well as power consumption. Results have shown that different data exchange protocols can have a significant impact on a decision support system response time.

Why should you model time when you use Markov models for heart sound analysis

Auscultation is a widely used technique in clinical activity to diagnose heart diseases. However, heart sounds are difficult to interpret because a) of events with very short temporal onset between them (tens of milliseconds) and b) dominant frequencies that are out of the human audible spectrum. In this paper, we propose a model to segment heart sounds using a semi-hidden Markov model instead of a hidden Markov model. Our model in difference from the state-of-the-art hidden Markov models takes in account the temporal constraints that exist in heart cycles. We experimentally confirm that semi-hidden Markov models are able to recreate the "true" continuous state sequence more accurately than hidden Markov models. We achieved a mean error rate per sample of 0.23.

A multi-spot exploration of the topological structures of the reconstructed phase-space for the detection of cardiac murmurs

Acoustic heart signals are generated by a turbulence effect created when the heart valves snap shut, and therefore carrying significant information of the underlying functionality of the cardiovascular system. In this paper, we present a method for heart murmur classification divided into three major steps: a) features are extracted from the heart sound; b) features are selected using a Backward Feature Selection algorithm; c) signals are classified using a K-nearest neighbor's classifier. A new set of fractal features are proposed, which are based on the distinct signatures of complexity and self-similarity registered on the normal and pathogenic cases. The experimental results show that fractal features are the most capable of describing the non-linear structure and the underlying dynamics of heart sounds among the all feature families tested. The classification results achieved for the mitral auscultation spot (88% of accuracy) are in agreement with the current state of the art methods for heart murmur classification.

Comparison between users of a new methodology for heart sound auscultation

Auscultation is a routine exam and the first line of screening in heart pathologies. The objective of this study was to assess if using a new data collection system, the DigiScope Collector, with a guided and automatic annotation of heart auscultation, different levels of expertise/experience users could collect similar digital auscultations. Data were collected within the Heart Caravan Initiative (Paraíba, Brasil). Patients were divided into two study groups: Group 1 evaluated by a third year medical student (User 1), and an experienced nurse (User 2); Group 2 evaluated by User 2 and an Information Technology professional (User 3). Patients were auscultated sequentially by the two users, according to the randomization. Features extracted from each data set included the length (HR) of the audio files, the number of repetitions per auscultation area, heart rate, first (S1) and second (S2) heart sound amplitudes, S2/S1, and aortic (A2) and pulmonary (P2) components of the second heart sound and relative amplitudes (P2/A2). Features extracted were compared between users using paired-sample test Wilcoxon test, and Spearman correlations (P<;0.05 considered significant). Twenty-seven patients were included in the study (13 Group 1, and 14 Group 2). No statistical significant differences were found between groups, except in the time of auscultation (User 2 consistently presented longer auscultation time). Correlation analysis showed significant correlations between extracted features from both groups: S2/S1 in Group 1, and S1, S2, A2, P2, P2/A2 amplitudes, and HR in Group 2. Using the DigiScope Collector, we were able to collect similar digital auscultations, according to the features evaluated. This may indicate that in sites with limited access to specialized clinical care, auscultation files may be acquired and used in telemedicine for an expert evaluation.

Exploring embedding matrices and the entropy gradient for the segmentation of heart sounds in real noisy environments

In this paper we explore a novel feature for the segmentation of heart sounds: the entropy gradient. We are motivated by the fact that auscultations in real environments are highly contaminated by noise and results reinforce our suspicions that the entropy gradient is not only robust to such noise but maintains a high sensitivity to the S1 and S2 components of the signal. Our whole approach consists of three stages, out of which the last two are novel contributions to this field. The first stage consists of typical pre-processing and wavelet reconstruction to obtain the Shannon energy envelogram. On the second stage we use an embedding matrix to track the dynamics of the system, which is formed by delay vectors with higher dimension than the corresponding attractor. On the third stage, we use the eigenvalues and eigenvectors of the embedding matrix to estimate the entropy of the envelogram. Finite differences are used to estimate entropy gradients, in which standard peak picking approaches are used for heart sound segmentation. Experiments are performed on a public dataset of pediatric auscultations obtained in real environments and results show the promising potential of this novel feature for such noisy scenarios.

Vital analysis: field validation of a framework for annotating biological signals of first responders in action

First responders are professionals that are exposed to extreme stress and fatigue during extended periods of time. That is why it is necessary to research and develop technological solutions based on wearable sensors that can continuously monitor the health of these professionals in action, namely their stress and fatigue levels. In this paper we present the Vital Analysis smartphone-based framework, integrated into the broader Vital Responder project, that allows the annotation and contextualization of the signals collected during real action. After a contextual study we have implemented and deployed this framework in a firefighter team with 5 elements, from where we have collected over 3300 hours of annotations during 174 days, covering 382 different events. Results are analysed and discussed, validating the framework as a useful and usable tool for annotating biological signals of first responders in action.