Method and apparatus for determining a heart period from an ECG waveform using image representation of ECG

Abstract

A method and system and computer program product for estimating a heart period is disclosed. The heart period is detected from an ECG recording. ECG data is acquired, and converted into electronic ECG images. The data is processed to prepare for estimation of a heart period. The heart period is estimated based upon an average of intervals between a plurality of detected peaks of electronic electrocardiogram waveforms. The peaks are determined by taking a product of a filtered electronic ECG signal with a wandering baseline removed, a difference between the upper and lower ECG envelopes of the electronic ECG images, and a first order derivative of a derived ECG waveform.

Description

BACKGROUND OF THE INVENTION[0001]The present invention relates generally to the field of electrocardiograms (ECG), and more specifically, automatic processing of electrocardiograms.[0002]There has been an increased need for automatic processing of the electrocardiogram (ECG). Many systems have been developed to perform signal processing tasks such as 12-lead off-line ECG analysis, and real-time patient monitoring. These applications require an accurate detection of the heart rate by the ECG.[0003]Although many hospitals now have digital ECG records, much of the ECG data is in paper form. Unlocking these ECG records printed on paper and exposing them to digital analysis would be useful.[0004]Digital ECG recordings are sampled finely, and contain noisy data with baseline wandering problems. The heart beats that illustrate a disease may be carefully screened by technicians before taking the recording, so that the paper version may show only a few seconds of data needed for diagnosis.[0...

AI Technical Summary

Benefits of technology

[0007]In one aspect of the invention, a method for estimating a heart period using at least one of a plurality of electrocardiogram recordings is disclosed. The method comprises utilizing a computer processor to obtain the plurality of electrocardiogram recordings generated from a plurality of electrocardiogram sources; converting the plurality of the electrocardiogram recordings into a plurality of corresponding electronic electrocardiogram images; removing noise from the corresponding electronic electrocardiogram images; deriving an electrocardiogram waveform from one of the electronic electrocardiogram images by extracting an upper envelope and a lower envelope of the electrocardiogram waveform from the corresponding electronic electrocardiogram image using a curve tracing algorithm, the algorithm for extracting the upper envelope and lower envelope comprising: tracing upper and lower edges of a curve of the electrocardiogram waveform from one of the electronic electrocardiogram images; defining two search ranges centered around the upper and lower edges of the curve of the electrocardiogram waveform; using at least one of: a morphological open algorithm to fill in holes inside the curve of the electrocardiogram waveform and a morphological close algorithm to eliminate noise pixels near the curve of the electrocardiogram waveform; closing small gaps by closing the electrocardiogram waveform between inserted seed points along an expected location of the electrocardiogram waveform; and closing large gaps by inserting a segment tangent to the curve of the electrocardiogram waveform; determining an electrocardiogram signal peak from increased thickness at the electrocardiogram signal peak due to pixel dithering; determining a difference between the upper envelope and the lower envelope at the electrocardiogram signal peak and at an electrocardiogram low point, which represent a local maximum and a local minimum of one of the corresponding electronic electrocardiogram images; estimating the heart period by: computing a first order derivative of the derived electrocardiogram waveform for amplifying high frequencies of the derived electrocardiogram waveform; determining a peak of the derived electrocardiogram waveform by taking a product of a filtered electrocardiogram signal from one of the corresponding electronic electrocardiogram images, with a wandering baseline removed, a difference between the upper envelope and the lower envelope of the derived electrocardiogram waveform from each of the at least one of the corresponding electronic electrocardiogram images and the first order derivative of the derived electrocardiogram waveform; and determining an average of intervals between a plurality of detected peaks from at least one region selected from the derived electrocardiogram waveform; and outputting the determined heart period on a user interface display.

Problems solved by technology

Digital ECG recordings are sampled finely, and contain noisy data with baseline wandering problems.

While it is difficult to estimate the heart cycle directly from a depicted heart region in a video, it is relatively easy to estimate the heart rate from a synchronizing ECG.

Images