Segmentation of heart sound recordings by a duration-dependent hidden Markov model

Section: Preparation and Analysis Of The Signalsmentioning

confidence: 99%

PCG Classification Using a Neural Network Approach

Grzegorczyk

Soliński

Łepek

et al. 2016

“…For this purpose the algorithm provided in the sample entry was used with minor modifications [3]. We calculated the cardiac cycle lengths and the systolic lengths for the training database by using the hand-corrected annotation of the recordings.…”

Section: Preprocessingmentioning

confidence: 99%

Morphological Determination of Pathological PCG signals by Time and Frequency Domain Analysis

Goda

Hajas²

2016

With the development of diagnostic devices over the past few decades, the algorithmic classification of heart sound recordings has become possible. Although this field has been under research for a relatively long time, the classification of such recordings is not yet straightforward.We were given a large manually classified database of heart sounds with the challenge [1]. We worked together with an experienced cardiologist to find the aspects affecting the classifications.To algorithmically classify a heart sound recording as normal or abnormal, it is necessary in most cases to accurately locate both the fundamental heart sounds and the systolic and diastolic regions. For this purpose we used the method provided in the example entry [2][3]. Minor modifications were made, such as tuning some of the parameters to match the database parameters.In the classification of the heart sounds, we were looking for the morphological features of the abnormal signals, for example, mitral stenosis, mitral insufficiency, aortic stenosis, aortic insufficiency, tricuspid stenosis and tricuspid insufficiency. We extracted several features from both time and frequency domains, for example, the frequency properties of systolic and diastolic segments and resampled wavelet envelope features.The extracted features were classified by the help of a support vector machine. In order to train the classifier, we used a reduced, sorted dataset with a more balanced ratio of abnormal and normal signals. During the official phase, our best scores on a random subset were 77.2% sensitivity, 85.2% specificity and 81.2% final modified accuracy (MAcc). Our scores for the entire test dataset are 83.77% sensitivity, 76.8% specificity and 80.28% MAcc. Our scores for the entire training dataset are 93.08% sensitivity, 84.70% specificity and 88.70% MAcc.

“…Hence signals were filtered with a band-pass Butterworth filter of the frequency range, 25 Hz to 400 Hz to remove high-frequency noise as well as artifacts such as baseline wandering. The signal spikes were then removed using Schmidt spike removal technique [4] and the signal was normalized to zero mean and unit variance. Reference annotations for four heart sound states (S1, systole, S2, diastole), for each heart beat, were then obtained for the pre-processed signals using Springer's segmentation algorithm [5] which is a state of the art solution for heart beat segmentation.…”

Section: Preprocessingmentioning

confidence: 99%

A Novel Approach for Classification of Normal/Abnormal Phonocardiogram Recordings using Temporal Signal Analysis and Machine Learning

Vernekar

Nair²,

Vijaysenan

et al. 2016

This paper discusses a novel approach used for classification of phonocardiogram (PCG) excerpts into normal and abnormal classes as a part of Physionet 2016 challenge [10]. The dataset used for the competition comprises of cardiac abnormalities such as mitral valve prolapse (MVP), benign murmurs, aortic diseases, coronary artery disease, miscellaneous pathological conditions etc.[3], We present the approach used for classification from a general machine learning application standpoint, giving details on feature extraction, type of classifiers used comparing their performances individually and in combination. We propose a technique which leverages previous research on feature extraction with a novel approach to modeling temporal dynamics of the signal using Markov chain analysis [7, 9]. These newly introduced Markov features along with other statistical and frequency domain features, trained over an ensemble of artificial neural networks and gradient boosting trees, with bagging, gave us an accuracy of 82% on the validation dataset provided in the competition and was consistent with the test data with the best result of 78%. IntroductionThis work describes a novel approach designed for Physionet 2016 Challenge Classification of Normal/Abnormal Heart Sound Recordings. The objective here is to classify Phonocardiogram (PCG) recordings into normal and abnormal categories. A comprehensive detail on the database, explaining how the PCG signals were collected and the type of abnormalities found are discussed in the paper mentioned in the reference section [3].Usually, statistical features such as means, standard deviations of systole, diastole intervals, and signal complexity features are used as features for classification and these are enough to give decent results. But these features fail to completely capture the temporal information of the signal. This could be very important since it represents how each heart beat changes over time.To capture the temporal dynamics, we take PCG signal beat by beat and assign each beat a symbol/category based on different thresholds set on features (ratios of systole intervals to RR interval, diastole intervals to RR interval, beat energy, the power of frequency component above 200 Hz). Thus, a sequence of symbols for the entire signal is obtained. We then extract features out of this sequence for classification. One of the ways we employed is to create a Markov chain with symbols being the states of the matrix and the resulting transition probabilities are used as features. These features along with marginal probabilities of states and rest of the acoustic features like sample entropy, instantaneous frequency analysis etc. are used to train an ensemble/bag of 4 class boosted tree classifiers [8, 13] and 4 artificial neural networks [12]. PreprocessingThe training data consists of PCG signals of varying length, anywhere between 5s to just over 120s all sampled at 2000 Hz. For training, all the signals were re-sampled to 1000 Hz and features were extracted. Since PCG record...