System and method for detecting a characteristic in an ECG waveform

Patent metadata

Field	Value
Publication number	US-9724007-B2
Application number	US-201214426967-A
Country	US
Kind code	B2
Filing date	Sep 11, 2012
Priority date	Sep 11, 2012
Publication date	Aug 8, 2017
Grant date	Aug 8, 2017

Abstract

An apparatus and method is provided that identifies the presence or absence of a P-wave within a set of ECG data. A computation processor identifies the R-wave and then analyses a section of the waveform within a predetermined time window preceding the detected R-wave peak. The waveform within the window is analysed to identify a candidate P-wave, and in response to identifying the candidate P-wave a first and second feature associated therewith is measured. A composite feature value is calculated from the first and second measures, and compared to a classification threshold value. In an exemplary embodiment, the first feature represents a height between a highest peak of the candidate P-wave and a trough of the Q-wave, and the second feature represents a time between the peak of the candidate P-wave and a peak of the R-wave.

First claim

What is claimed is:

1. An apparatus that detects at least one characteristic associated with an input signal waveform representing an electrocardiograph (ECG) signal of a patient, the at least one characteristic being detected by the apparatus is the presence of a P-wave, the apparatus comprising:
an acquisition processor that acquires data representing a sequence of successive pulses of heartbeat of a patient;
a computation processor electrically coupled to said acquisition processor that, for each heartbeat pulse,
identifies a portion of the input signal that may include the at least one characteristic and, for the identified portion of the input signal, measures a first and second feature associated therewith the identified portion of the input signal representing a candidate P-wave, wherein the portion of the input signal identified by the computation processor is within a window of time occurring at a predetermined time prior to an R-wave of a same heartbeat pulse;
in response to detecting a plurality of peaks within the window, selects one of the plurality of peaks as candidate data representative of the presence of the at least one characteristic;
selects a first peak of the plurality of peaks having the greatest amplitude;
measures a distance between the first peak and a peak of an R-wave of a preceding heartbeat waveform;
compares the measured distance to a predetermined threshold, wherein the selected first peak is identified as the candidate data when the measured distance exceeds the predetermined threshold;
calculates a composite feature value using the first and the second feature, the composite feature value representing the likelihood that the at least one characteristic is present in the identified portion of the input signal;
compares the calculated composite feature value to a classification threshold value; and

an output processor electrically coupled to said computation processor that generates data indicating the presence of the P-wave in response to a determination that the calculated composite feature value exceeds the classification threshold.

2. The apparatus according to claim 1 , wherein
the computation processor selects as candidate data representative of the at least one characteristic, a peak of the input signal waveform within the window having a greatest amplitude and measures the first and second features of the candidate data of the peak.

3. The apparatus according to claim 1 , wherein
in response to determining that the measured distance is less than the predetermined threshold, the computation processor selects a peak within a window having a next highest amplitude and occurring subsequent to the first peak; measures a distance between the peak having the next highest amplitude with the predetermined threshold, wherein the peak having the next highest amplitude is identified as the candidate data when the measured distance exceeds the predetermined threshold.

4. The apparatus according to claim 1 , wherein
the first feature measured by the computation processor represents a height of a peak of the identified portion of the waveform and a trough of a subsequently occurring wave of the heartbeat waveform, and
the second feature measured by the computation processor represents a time between a peak of the identified portion of the waveform and a peak of a second subsequently occurring wave of the heartbeat waveform.

5. The apparatus according to claim 1 , wherein the computation processor
measures the first feature by determining a height of a peak of the candidate P-wave above a trough of a Q-wave of a same heartbeat, and
measures the second feature by determining a time between a peak of the candidate P-wave and a peak of the R-wave of the respective heartbeat.

6. The apparatus according to claim 1 , wherein
the computation processor calculates the composite feature value by multiplying data representing the first feature by a first weight value of a weighting vector and multiplies data representing the second feature by a second weight value of the weighting vector.

7. The apparatus according to claim 6 , wherein
the weighting vector includes weight values that correspond to respective strengths associated with the ability of each of the first and second features to successfully identify the presence of the at least one characteristic in the identified portion of the waveform.

8. The apparatus according to claim 1 , wherein
the computation processor selects the classification threshold value by classifying a pattern within a set of training data that includes a heartbeat waveform having the at least one characteristic.

9. The apparatus according to claim 1 , wherein
the computation processor identifies a further characteristic associated with the input signal, and uses the presence of the P-waves in the respective heartbeats of the successive heartbeats in conjunction with the identified further characteristic to determine if the patient is experiencing a cardiac condition.

10. The apparatus according to claim 9 , wherein
the further characteristic is a pulse irregularity measurement of heartbeats in the successive pulses of heartbeat data, wherein the computation processor calculates a pulse interval irregularity measure based on a sum of time interval differences occurring between pairs of successive pulses of said sequence of successive pulses and a mean time interval of successive pulses of said sequence and excluding time interval differences exceeding a predetermined maximum threshold from said sum.

11. The apparatus according to claim 10 , wherein the computation processor
calculates a P-wave presence indicator by determining a percentage of heartbeats in the successive heartbeats having P-waves, and
compares the determined pulse irregularity value to a first threshold value and the P-wave presence indicator to a second threshold value, the cardiac condition is determined to be present if the pulse irregularity value exceeds the first threshold and the P-wave presence indicator is below the second threshold value.

12. A method of detecting at least one characteristic associated with an input signal waveform representing an electrocardiogram (ECG) signal of a patient, the at least one characteristic being detected is the presence of a P-wave, the method comprising the activities of:
acquiring, by an acquisition processor, data representing a sequence of successive pulses of heartbeat of a patient;
identifying, by a computation processor, for each heartbeat, a portion of the input signal that may include the at least one characteristic, the identified portion of the input signal representing a candidate P-wave, wherein the portion of the input signal identified by the computation processor is within a window of time occurring at a predetermined time prior to an R-wave of a same heartbeat pulse;
selecting one of the plurality of peaks as candidate data representative of the at least one characteristic in response to detecting a plurality of peaks within the window;
selecting a first peak of the plurality of peaks having the greatest amplitude;
measuring a distance between the first peak and a peak of an R-wave of a preceding heartbeat waveform;
comparing the measured distance to a predetermined threshold;
identifying the selected first peak as the candidate data when the measured distance exceeds the predetermined threshold;
measuring a first and second feature associated with the identified portion of the input signal;
calculating a composite feature value using the first and the second feature, the composite feature value representing a likelihood that the at least one characteristic is presen

Assignees

• Cole Bryan Thomas
• Draeger Medical Systems Inc

Inventors

• Cole Bryan Thomas

Classifications

• A61B5/353 Primary

  Detecting P-waves · CPC title

• A61B5/04012 Primary

  Human Necessities · mapped topic

• A61B5/0472

  Human Necessities · mapped topic

• A61B5/7246 Primary

  using correlation, e.g. template matching or determination of similarity · CPC title

• A61B5/0456

  Human Necessities · mapped topic