SYSTEM AND METHOD FOR CLASSIFYING A HEART SOUND

• US 20130226019A1
• Filed: 08/25/2011
• Published: 08/29/2013
• Est. Priority Date: 08/25/2010
• Status: Active Grant

• Current Assignee: Diacoustic Medical Devices (Pty) Ltd.
• Original Assignee: Diacoustic Medical Devices (Pty) Ltd.
• Inventors: Pretorius, Eugene, Cronje, Matthys Louwrens, Strydom, Otto Marsl
• Granted Patent: US 8,992,435 B2
• Time in Patent Office: Days
• US Class Current: 600/528
• CPC Class Codes: A61B 5/02 Detecting, measuring or recording pulse, heart rate, blood pressure or blood flow; Combined pulse/heart-rate/blood pressure determination; Evaluating a cardiovascular condition not otherwise provided for, e.g. using combinations of techniques provided for in this group with electrocardiography or electroauscultation; Heart catheters for measuring blood pressure
• CPC Class Codes: A61B 5/349 Detecting specific parameters of the electrocardiograph cycle
• CPC Class Codes: A61B 5/352 Detecting R peaks, e.g. for synchronising diagnostic apparatus; Estimating R-R interval
• CPC Class Codes: A61B 5/7264 Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems neural networks per se G06N3/00; expert systems per se G06N5/00
• CPC Class Codes: A61B 5/7267 involving training the classification device
• CPC Class Codes: A61B 7/04 Electric stethoscopes microphones, acoustic transducers therefor H04R

Abstract

A method and system for electronically classifying a pre-processed heart sound signal of a patient as functional (normal) or pathological is provided. The pre-processed patient heart sound signal is segmentised and features are extracted therefrom (104) to build up a feature vector which is representative of the heart sound signal. The feature vector is then fed to a diagnostic decision support network (105) comprising a plurality of artificial neural networks, each relating to a known heart pathology, which is in turn used to conduct the classification.

Claims

1. A method of classifying a pre-processed heart sound signal of a patient, the method being characterised in that the pre-processed patient heart sound signal is electronically compared to a plurality of similarly pre-processed heart sounds corresponding to each of a plurality of pathologies and a diagnosis made by determining whether the patient heart sound corresponds to the heart sounds of any one or more of the pathologies.

2. A method as claimed in claim 1 wherein determining whether the patient heart sound corresponds to the heart sounds of any one or more of the pathologies includes determining a confidence measure to indicate the confidence with which a diagnosis is made.

3. A method as claimed in claim 1 wherein an artificial neural network (ANN) is provided for each pathology, each artificial neural network being trained with at least features extracted from the pre-processed heart sounds corresponding to the applicable pathology, and wherein each artificial neural network is used to determine whether the pre-processed patient heart sound corresponds to the pathology associated therewith.

4. A method as claimed in claim 1 wherein a diagnostic decision support network is provided which comprises the artificial neural networks for each pathology, the diagnostic decision support network being used to determine if the pre-processed patient heart sound signal corresponds to the heart sounds of any one or more of the pathologies.

5. A method as claimed in claim 3 in which the pre-processed heart sound signal is segmentised and features are extracted therefrom, the features being compiled into a feature vector which serves as input to the diagnostic decision network.

6. A method as claimed in claim 1 in which the heart sound signal is pre-processed by identifying or predicting locations of individual heart beats within the heart sound signal, predicting locations and durations for systole and diastole segments of the heart sound signal, compiling a peak array containing peak locations within the heart sound signal that correspond to possible locations of S1 and S2 heart sound pulses in the heart sound signal, calculating interval sizes between the peak locations in the peak array, selecting and isolating heart beats in the heart sound signal that were detected with the highest level of confidence based on the predicted durations of the systole and diastole segments of the heart sound signal and the peak array, determining if segmentation of the heart sound signal is possible based on the selected and isolated heart beats, dividing inter-peak intervals that make up the heart sound signal into even and odd groups in the order in which they appear if segmentation of the heart sound signal was found to be possible, and calculating the variance and mean for both groups and taking the group with the largest ratio of variance over mean as corresponding to the diastole intervals and the other group as corresponding to the systole intervals.

7. A system for classifying a pre-processed heart sound signal of a patient, the system including a processor configured to compare the pre-processed patient heart sound to a plurality of similarly pre-processed heart sounds corresponding to each of a plurality of pathologies and to provide a diagnosis by determining whether the patient heart sound signal corresponds to the heart sounds of any one or more of the pathologies.

8. A system as claimed in claim 7 wherein the processor is further configured to calculate a confidence measure indicative of the confidence with which a diagnosis is made.

9. A system as claimed in claim 7 which includes a diagnostic decision support network comprising an artificial neural network (ANN) associated with each pathology, the artificial neural networks being trained with at least features extracted from the pre-processed heart sounds associated with its pathology and the processor being further configured to use the diagnostic decision support network to provide the diagnosis.

10. A system as claimed in claim 7 wherein the processor is configured to segmentise the heart sound signal and to extract features therefrom, and to compile the features into a feature vector which serves as input to the diagnostic decision support network.