Apparatus and method for discriminating between cardiac rhythms on the basis of their morphology using a neural network

• US 5,542,430 A
• Filed: 09/16/1994
• Issued: 08/06/1996
• Est. Priority Date: 09/16/1994
• Status: Expired due to Fees

• Current Assignee: Pacesetter Incorporated (Abbott Laboratories Incorporated)
• Original Assignee: Telectronics Pacing Systems, Inc (Abbott Laboratories Incorporated)
• Inventors: Farrugia, Steven, Yee, Hansen, Nickolls, Peter
• Primary Examiner(s): Kamm, William E.
• Application Number:
  US08/306,993
• Time in Patent Office: 690 Days
• Field of Search: 128/702-705
• US Class Current: 600/518
• CPC Class Codes: A61B 5/363 Detecting tachycardia or bradycardia
• CPC Class Codes: A61B 5/7242 using integration
• 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: G16H 50/20 for computer-aided diagnosis, e.g. based on medical expert systems
• CPC Class Codes: Y10S 128/924 using artificial intelligence
• CPC Class Codes: Y10S 128/925 Neural network

Abstract

An apparatus and a method are provided for coupling to a patient'"'"'s heart for discriminating between tachycardias of physiological origin, and those of pathological origin having similar rates; and also for discriminating amongst those of pathological origin having similar rates. The apparatus includes transducers and/or sensing electrodes in either or both the atrium and/or ventricle. Also included are signal processing elements for determining the times of atrial and ventricular events and for extracting morphological features from the waveforms, and a neural network for classifying the heart rhythm. The method includes a step of discriminating between different types of heart rhythms having overlapping rates. The method utilizes atrial-atrial, ventricular-ventricular and atrio-ventricular intervals; integrated waveforms; sums of differences of waveform samples; rectified integrated bandpass filtered waveforms; numbers of zero crossings in the electrogram; area under the ventricular electrogram; and R wave slope, QR area and RS area of the electrogram.

Claims

1. A device for monitoring and classifying cardiac rhythms of a patient'"'"'s heart, comprising:

means for coupling to at least one of an atrium and a ventricle of the heart, for sensing respective ones of an atrial and a ventricular thereof as waveforms which are characteristic of respective cardiac rhythms;

feature extraction means coupled to said coupling means and responsive to the sensed waveforms, for extracting from said waveforms features relating to the timing of events and the morphology of the waveforms;

neural network means coupled to said feature extraction means for classifying said morphological features, said neural network means further including

first means for discriminating between features relating to rhythms of normal origin and rhythms of pathological origin, and

second means for discriminating between features relating to different kinds of tachycardia of pathological origin

said first and second discriminating means generating output signals classifying cardiac rhythms into physiological and pathological categories;

wherein said neural network means includes means for setting a time interval;

wherein said feature extraction means includes a threshold tracking peak detector means to provide at least one fiducial point in each sensed waveform, said fiducial point being one of a maximum positive and minimum negative peak within said time of the threshold being exceeded by the waveform;

wherein said feature extraction means integrates the waveform about its fiducial point for a predetermined period of time before and after said point.

2. A device according to claim 1 wherein:

said classifying means utilizes said first and second means to discriminate between different types of rhythms whose rates may overlap in frequency.

3. A device for monitoring and classifying cardiac rhythms of a patient'"'"'s heart, comprising:

means for coupling to at least one of an atrium and a ventricle of the heart, for sensing respective ones of an atrial and a ventricular waveforms which are characteristic of respective cardiac rhythms;

feature extraction means coupled to said coupling means and responsive to the sensed waveforms, for extracting from said waveforms features relating to the timing of events and the morphology of the waveforms;

neural network means coupled to said feature extraction means for classifying said morphological features, said neural network means further including

first means for discriminating between features relating to rhythms of normal origin and rhythms of pathological origin, and

second means for discriminating between features relating to different kinds of tachycardia of pathological origin

wherein said neural network means includes means for setting a time interval;

said first and second discriminating means generating output signals classifying cardiac rhythms into physiological and pathological categories;

wherein said feature extraction means includes a threshold tracking peak detector means to provide at least one fiducial point in each sensed waveform, said fiducial point being one of a maximum positive and minimum negative peak within a preselected time of the threshold being exceeded by the waveform;

wherein said feature extraction means takes sums of differences of samples taken at predetermined intervals from the waveform symmetrically about its fiducial point.

4. A device for monitoring and classifying cardiac rhythms of a patient'"'"'s heart, comprising:

means for coupling to at least one of an atrium and a ventricle of the heart, for sensing respective ones of an atrial and a ventricular waveforms which are characteristic of respective cardiac rhythms;

feature extraction means coupled to said coupling means and responsive to the sensed waveforms, for extracting from said waveforms features relating to the timing of events and the morphology of the waveforms;

neural network means coupled to said feature extraction means for classifying said morphological features, said neural network means further including

first means for discriminating between features relating to rhythms of normal origin and rhythms of pathological origin, and

second means for discriminating between features relating to different kinds of tachycardia of pathological origin

wherein said neural network means includes means for setting a time interval;

said first and second discriminating means generating output signals classifying cardiac rhythms into physiological and pathological categories;

wherein said feature extraction means includes a threshold tracking peak detector means to provide at least one fiducial point in each sensed waveform, said fiducial point being one of a maximum positive and minimum negative peak within a preselected time of the threshold being exceeded by the waveform;

wherein said feature extraction means further includes a plurality of bandpass filters and integrators for filtering the waveform through said filters and integrators during the interval between successive fiducial points.

5. A device for monitoring and classifying cardiac rhythms of a patient'"'"'s heart, comprising:

means for coupling to at least one of an atrium and a ventricle of the heart, for sensing respective one of an atrial and a ventricular thereof as waveforms which are characteristic of respective cardiac rhythms;

feature extraction means coupled to said coupling means and responsive to the sensed waveforms, for extracting from said waveforms features relating to the timing of events and the morphology of the waveforms;

neural network means coupled to said feature extraction means for classifying said morphological features, said neural network means further including

first means for discriminating between features relating to rhythms of normal origin and rhythms of pathological origin, and

second means for discriminating between features relating to different kinds of tachycardia of pathological origin;

wherein said neural network means includes means for setting a time interval;

said first and second discriminating means generating output signals classifying cardiac rhythms into physiological and pathological categories;

wherein said feature extraction means includes a threshold tracking peak detector means to provide at least one fiducial point in each sensed waveform, said fiducial point being one of a maximum positive and minimum negative peak within a preselected time of the threshold being exceeded by the waveform;

wherein said feature extraction means further includes means for determining the crossings of a waveform through zero amplitude for determining the crossings of the waveforms through zero amplitude.

6. A device for monitoring and classifying cardiac rhythms of a patient'"'"'s heart, comprising:

means for coupling to at least one of an atrium and a ventricle of the heart, for sensing respective ones of an atrial and a ventricular thereof as waveforms which are characteristic of respective cardiac rhythms;

feature extraction means coupled to said coupling means and responsive to the sensed waveforms, for extracting from said waveforms features relating to the timing of events and the morphology of the waveforms;

neural network means coupled to said feature extraction means for classifying said morphological features, said neural network means further including

first means for discriminating between features relating to rhythms of normal origin and rhythms of pathological origin, and

second means for discriminating between features relating to different kinds of tachycardia of pathological origin;

wherein said neural network means includes means for setting a time interval;

said first and second discriminating means generating output signals classifying cardiac rhythms into physiological and pathological categories;

wherein said feature extraction means includes a threshold tracking peak detector means to provide at least one fiducial point in each sensed waveform, said fiducial point being one of a maximum positive and minimum negative peak within a preselected time of the threshold being exceeded by the waveform;

wherein said feature extraction means determines an interval between successive atrial and ventricular complexes of the waveform; and

wherein said classifying means classifies as of pathological origin those of such tachycardias in which atrioventricular intervals change rapidly in sequence.