Back to results
US 8,135,456 B2Granted
Device and method for detecting atrial fibrillation
AssigneePacesetter, Inc.
InventorJeffrey A. Haluska
Life of this patent
Estimating the term…
Abstract
Detection of atrial fibrillation involves detecting a plurality of ventricular events and obtaining a series of probabilities of AF, each corresponding to a probability of AF for a different beat window having a plurality of ventricular events. AF onset is detected when one or each of a plurality of consecutive AF probabilities satisfies an AF trigger threshold. AF termination is detected when one or each of a plurality of consecutive AF probabilities does not satisfy the AF trigger threshold. A probability of AF is obtained using one or more of a probability of AF that is based on the presence of irregularity of the ventricular events (PMC); a probability of AF that is based on variances of R-R intervals of the ventricular events (PVAR); and/or a combination of these probabilities (PCOMBINED).
Claims10
A cardiac rhythm monitoring device comprising:
- electrodes adapted to be coupled to a patient to sense cardiac electrical activity;
- and
- a processor coupled to the electrodes to receive the electrical activity and programmed to detect ventricular events within the electrical activity, wherein a plurality of successive detected ventricular events define a first beat window;
- and for the first beat window:
- calculate RR intervals for each pair of adjacent ventricular events;
- calculate the differences (dRR) between successive RR intervals;
- determine a first probability of AF based on the presence of irregularity of the ventricular events, wherein to determine the first probability the programmer is further programmed to convert the dRR values into bin values based on a plurality of bin thresholds;
- calculate a transition measurement between successive bin values;
- calculate a sample chain of statistics using the transition measurements;
- compare the sample chain of statistics to a known AF chain of statistics and to a known non-AF chain of statistics;
- determine a measure of AF quality based on the comparisons;
- calculate a heart rate index based on the RR intervals;
- use a look-up table that correlates AF quality and heart rate index to a probability of AF to determine the first probability of AF;
- determine a second probability of AF based on variances of R-R intervals of the ventricular events, wherein to determine the second probability the programmer is further programmed to calculate a first variance over positive and negative dRR values;
- calculate a second variance only over the negative dRR values;
- calculate a third variance only over the positive dRR values;
- use a look-up table that correlates the first, second and third variances to a probability of AF to determine the second probability of AF;
- and
- combine the first and second probabilities of AF to obtain a first-beat-window combined probability of AF, wherein the combined probability of AF ranges between a minimum value and a maximum value;
- compare the combined probability of AF to an AF trigger threshold, wherein the AF trigger threshold is a predetermined value between the minimum value and the maximum value;
- and
- detect AF onset when the first-beat-window combined probability of AF is greater than or equal to the AF trigger threshold.
19 more elements
A method of detecting atrial fibrillation comprising:
- detecting electrical activity using electrodes adapted to be coupled to a patient;
- detecting ventricular events within electrical cardiac activity using a processor, wherein a plurality of successive detected ventricular events define a first beat window;
- for the first beat window, performing the following using the processor:
- calculating RR intervals for each pair of adjacent ventricular events;
- calculating the differences (dRR) between successive RR intervals;
- determining a first probability of AF based on the presence of irregularity of the ventricular events by:
- converting the dRR values into bin values based on a plurality of bin thresholds;
- calculating a transition measurement between successive bin values;
- calculating a sample chain of statistics using the transition measurements;
- comparing the sample chain of statistics to a known AF chain of statistics and to a known non-AF chain of statistics;
- determining a measure of AF quality based on the comparisons;
- calculating a heart rate index based on the RR intervals;
- using a look-up table that correlates AF quality and heart rate index to a probability of AF to determine the first probability of AF;
- determining a second probability of AF based on variances of R-R intervals of the ventricular events by:
- calculating a first variance over positive and negative dRR values;
- calculating a second variance only over the negative dRR values;
- calculating a third variance only over the positive dRR values;
- using a look-up table that correlates the first, second, and third variances to a probability of AF to determine the second probability of AF;
- and
- combining the first and second probabilities of AF to obtain a first-beat-window combine probability of AF, wherein the combined probability of AF ranges between a minimum value and a maximum value;
- and
- determining AF onset when the first-beat-window combined probability of AF is greater than or equal to the AF trigger threshold.
19 more elements
Dependencies are inferred from the claim text.