1,720,995 research outputs found
Multi-source signal processing in phonocardiography: comparison among signal selection and signal enhancement techniques*
No full textPhonocardiography (PCG) is a promising tool for the diagnosis and follow-up of cardiovascular diseases. To date it is available only in clinical settings, because it relies on an experienced examiner for the positioning of the electronic stethoscope. Making it possible for an unexperienced user to obtain high quality PCG signals would allow for developing instruments suitable for homecare purposes. In this work, we test the usage of three standardly positioned microphone probes. The aim is to compare two different approaches for enhancing the PCG signal quality, namely a) selecting the single source with the highest Signal-to-Noise Ratio (SNR) and b) combining the three sources through array signal processing techniques. Both approaches were tested on a sample population counting 24 healthy subjects. We found that the two approaches above give statistically different results (two-tailed paired t-test, p = 0.037) in terms of SNR of the enhanced signal. Specifically, selecting the single source with highest SNR gives, on average, the best results. Moreover, this approach is also associated with the lowest computational cost. Finally, for every subject of our sample population, we obtained SNR values higher than 12.5 dB on the enhanced signal, which we consider as sufficient for the application of heart sound segmentation and classification algorithms. We believe that this methodology allows for obtaining PCG signals of sufficient quality for the analysis of heart sounds, thus opening to the applicability of PCG in a homecare context
A Method for the Estimation of the Timing of Heart Sound Components Through Blind Source Separation in Multi-Source Phonocardiography
Full text linkRecently, phonocardiography (PCG) has gained importance as a diagnostic tool for cardiovascular diseases. In particular, the measurement of the time of occurrence of heart sounds may be of interest, in the clinical context, for the analysis of the electromechanical coupling of the heart. To date, though, there is no standardization concerning the positioning of the microphone probe over the chest, and this causes low accuracy and consistency in the measured timing values. Multi-source phonocardiography is a promising approach to face the stated issue. In this work, we present a methodology to estimate the latency of the components of the two main heart sounds towards
the corresponding R-wave peak based on the Blind Source Separation (BSS) of the contributions of the left and right side of the heart. We tested our algorithm on a sample population of 12 subjects over 10-minute long recordings of three simultaneous PCG signals and one electrocardiographic (ECG) signal for reference. Results show that the approach is robust with respect to the usage of different algorithms to perform BSS (FastICA, JADE). The measured timing values are consistent with what measured by means of a single-source algorithm we previously developed. This methodology looks promising in terms of obtaining accurate measurements of the time of occurrence of heart sound components and may have an impact in the clinical context
Personalized Detection of Motion Artifacts for Telemonitoring Applications
Full text linkAmong its main benefits, telemonitoring enables personalized
management of chronic diseases by means of biomarkers extracted from signals. In
these applications, a thorough quality assessment is required to ensure the reliability
of the monitored parameters. Motion artifacts are a common problem in recordings
with wearable devices. In this work, we propose a fully automated and personalized
method to detect motion artifacts in multimodal recordings devoted to the
monitoring of the Cardiac Time Intervals (CTIs). The detection of motion artifacts
was carried out by using template matching with a personalized template. The
method yielded a balanced accuracy of 86%. Moreover, it proved effective to
decrease the variability of the estimated CTIs by at least 17%. Our preliminary
results show that personalized detection of motion artifacts improves the robustness
of the assessment CTIs and opens to the use in wearable systems
Pulmonary Hypertension Detection from Heart Sound Analysis
Full text linkThe detection of Pulmonary Hypertension (PH) from the computer analysis of digitized heart sounds is a low-cost and non-invasive solution for early PH detection and screening. We present an extensive cross-domain evaluation methodology with varying animals (humans and porcine animals) and varying auscultation technologies (phonocardiography and seisomocardiography) evaluated across four methods. We introduce PH-ELM, a resource-efficient PH detection model based on the extreme learning machine that is smaller (300× fewer parameters), energy efficient (532× fewer watts of power), faster (36× faster to train, 44× faster at inference), and more accurate on out-of-distribution testing (improves median accuracy by 0.09 area under the ROC curve (auROC)) in comparison to a previously best performing deep network. We make four observations from our analysis: (a) digital auscultation is a promising technology for the detection of pulmonary hypertension; (b) seismocardiography (SCG) signals and phonocardiography (PCG) signals are interchangeable to train PH detectors; (c) porcine heart sounds in the training data can be used to evaluate PH from human heart sounds (the PH-ELM model preserves 88 to 95% of the best in-distribution baseline performance); (d) predictive performance of PH detection can be mostly preserved with as few as 10 heartbeats and capturing up to approximately 200 heartbeats per subject can improve performance
Comparison of different similarity measures in hierarchical clustering
Full text linkThe management of datasets containing heterogeneous types of data is a crucial point in the context of precision medicine, where genetic, environmental, and life-style information of each individual has to be analyzed simultaneously. Clustering represents a powerful method, used in data mining, for extracting new useful knowledge from unlabeled datasets. Clustering methods are essentially distance-based, since they measure the similarity (or the distance) between two elements or one element and the cluster centroid. However, the selection of the distance metric is not a trivial task: it could influence the clustering results and, thus, the extracted information. In this study we analyze the impact of four similarity measures (Manhattan or L1 distance, Euclidean or L2 distance, Chebyshev or L∞ distance and Gower distance) on the clustering results obtained for datasets containing different types of variables. We applied hierarchical clustering combined with an automatic cut point selection method to six datasets publicly available on the UCI Repository. Four different clusterizations were obtained for every dataset (one for each distance) and were analyzed in terms of number of clusters, number of elements in each cluster, and cluster centroids. Our results showed that changing the distance metric produces substantial modifications in the obtained clusters. This behavior is particularly evident for datasets containing heterogeneous variables. Thus, the choice of the distance measure should not be done a-priori but evaluated according to the set of data to be analyzed and the task to be accomplished
Comparison of Hierarchical and Partitional Clustering in Multi-Source Phonocardiography
Full text linkPhonocardiography (PCG) has proved a valuable tool over the years to monitor the status of at-risk patients for some cardiovascular diseases. Its multi-source version, consisting of the simultaneous recording of multiple acoustic signals from different points of the patient’s chest, is currently under research as a solution to develop wearable devices based on PCG and bring PCG to the patient’s domicile. When a high number of PCG signals are available, to define the most suitable auscultation area, depending on the clinical question, clustering comes into the picture. In this work, we applied agglomerative hierarchical clustering and k-means to multi-source PCG recordings. A similarity metrics based on the correlation of the signals was used to compare the signals based on their morphological characteristics. The two clustering methods resulted in a Rand Index averagely higher than 0.84, showing a high level of agreement and validating the usage of clustering for the application of interest. Hierarchical clustering allowed for obtaining a better trade-off between the intra-cluster variability and the inter-cluster distance. Adding to its deterministic nature, it should be considered as preferrable with respect to k-means. This work moves one step further to the development a reliable wearable device based on digital auscultation for the monitoring of the patient at its domicile
Stratification of Heart Sounds Morphology Through Unsupervised Learning
Full text linkThe use of heart sounds for the assessment of the hemodynamic condition of the heart in telemonitoring applications is object of wide research at date. Many different approaches have been tried out for the analysis of the first (S1) and second (S2) heart sounds, but their morphological interpretation is still to be explored: in fact, the sound morphology is not unique and this impact the separability of the heart sounds components with methods based on envelopes or model optimization. In this study, we propose a method to stratify S1 and S2 according to their morphology to explore their diversity and increase their morphological interpretability. The method we propose is based on unsupervised learning, which we obtain using the cascade of four Self-Organizing Maps (SOMs) of decreasing dimensions. When tested on a publicly available heart sounds dataset, the proposed clustering approach proved to be robust and consistent, with over 80% of the heartbeats of the same patient being clustered together. The identified heart sounds templates highlight differences in the time and energy domains which may open to new directions of analysis in the future
Can Multi-source Phonocardiography Enable Inexperienced Users to Record Heart Sounds for Telemonitoring Applications? A Comparative Analysis
Full text linkThe use of heart sounds in telemonitoring is extremely appealing due to portability, low cost, non-invasiveness. Nevertheless, the positioning of the electronic stethoscope is critical and prevents the use of phonocardiography (PCG) by inexperienced users. Multi-source PCG may provide a solution: by recording multiple signals from different points on the chest at high spatial resolution the problem of finding the best auscultation point is moved from the recording phase to the processing phase. In this study, we compare the quality of PCG signals recorded by inexperienced users through multi-source PCG against the quality of PCG signals recorded by an expert through a traditional single-source system. We enrolled 42 inexperienced volunteers and asked them to record signals on each other using a multi-sensor array that we designed for this purpose. An expert user also recorded signals from each volunteer from the four typical auscultation areas. Experimental results show that the multi-source system enabled inexperienced users to record signals of equal or better quality. We believe that our results lay the foundations to apply multi-source PCG in homecare
Separation of the Valvular Contribution to Heart Sounds through Blind Source Separation in Multi-Channel Phonocardiography
Full text linkThe separation of the contribution of the left and right cardiac valves to heart sounds is an open challenge in the field of phonocardiography. Yet, reliably measuring their time of closure in a noninvasive fashion would open to novel monitoring possibilities. In this work, we explore the potentiality of Blind Source Separation applied to multi-channel recordings at high spatial resolution to separate the components of the two main heart sounds. Our pipeline involves a pre-processing stage to isolate the segments of interest, a dimensionality reduction stage performed via clustering, and the application of Independent Component Analysis. Our results on a sample population of 52 healthy volunteers show a successful separation of the components. The estimated time of closure is consistent with the physiology of the heart sounds, and the statistical difference between the contributes of the valves from the same sound was proved. We believe that this work makes a step further towards the clinical use of heart sound components and lays the foundation to novel possibilities of analysis
Usability-driven design of medical device APPs for telemonitoring of chronic patients
Full text linkTelemonitoring systems, including devices, APPs,
and Medical Device Software, provide an irreplaceable
technological solution for the prevention of acute episodes in
chronic pathologies such as heart failure. Nevertheless, the
benefits are often compromised by complex designs which
hinder their adoption and their seamless integration into the
patient’s daily life. In this work, we propose a user-centered
design methodology to translate the specific requirements of
chronic patients into suitable technical specifications for
telemonitoring APP
- …
