Method and system to detect premature ventricular contractions in cardiac activity signals
A computer implemented method and system are provided for detecting premature ventricular contractions (PVCs) in cardiac activity. The method and system obtain cardiac activity (CA) signals for a series of beats, and, for at least a portion of the series of beats, calculate QRS scores for corresponding QRS complex segments from the CA signals. The method and system calculate a variability metric for QRS scores across the series of beats, calculate a QRS complex template using QRS segments from the series of beats, calculate correlation coefficients between the QRS complex template and the QRS complex segments, compare the variability metric to a variability threshold and the correlation coefficients to a correlation threshold, and designate the CA signals to include a predetermined level of PVC burden based on the comparing.
1. A computer implemented method for detecting premature ventricular contractions (PVCs) in cardiac activity, comprising:
under control of one or more processors configured with specific executable instructions,
obtaining cardiac activity (CA) signals for a series of beats;
for at least a portion of the series of beats, calculating QRS scores for corresponding QRS complex segments from the CA signals;
calculating a variability metric for QRS scores across the series of beats;
calculating a QRS complex template using QRS segments from the series of beats;
calculating correlation coefficients between the QRS complex template and the QRS complex segments,
comparing the variability metric to a variability threshold and the correlation coefficients to a correlation threshold, and
designating the CA signals to include a predetermined level of PVC burden based on the comparing.
2. The method of claim 1 , wherein the calculating the QRS scores comprises calculating at least one of a summation, area under a curve, or energy for the QRS complex segments.
3. The method of claim 1 , further comprising: creating a morphology ensemble from the QRS complex segments; and comparing the morphology ensemble to QRS morphologies for the corresponding QRS complex segments to obtain morphology correlation characteristics for the QRS complex segments.
4. The method of claim 1 , wherein the computing of the variability metric includes calculating a covariance Cov_QRS from the QRS scores, the covariance based on the QRS scores.
5. The method of claim 4 , wherein the covariance Cov_QRS represents a standard deviation of the QRS scores of the QRS complex segments divided by an average of the QRS scores, the QRS scores representing at least one of amplitude sums, energy or area under the curve for the corresponding QRS complex segments.
6. The method of claim 1 , wherein, when the variability metric does not satisfy the variability threshold, declaring the CA signals to include an non-significant PVC burden which is less than the predetermined level of PVC burden.
7. The method of claim 1 , wherein the morphology ensemble represents an ensemble average of the QRS morphologies for a desired number of the QRS complex segments; and the comparing includes comparing the QRS morphologies for the corresponding QRS complex segments to the ensemble average to obtain the morphology correlation characteristics.
8. The method of claim 7 , wherein the morphology correlation characteristics represent correlations between the ensemble average and the QRS morphologies of each of the QRS complex segments.
9. The method of claim 1 , further comprising rejecting the beats from the CA signals that exhibit a predetermined level of baseline drift.
10. The method of claim 1 , when the variability metric satisfies the variability threshold and the correlation coefficients satisfies the correlation conditions, declaring the CA signals to have significant PVC burden.
11. A system for detecting premature ventricular contractions (PVCs) in cardiac activity, comprising:
memory to store executable instructions;
one or more processors that, when executing the executable instructions, configured to:
obtain a cardiac activity (CA) signals for a series of beats;
for at least a portion of the series of beats, calculate QRS scores for corresponding QRS complex segments from the CA signals;
calculate a variability metric for QRS scores across the series of beats;
calculate a QRS complex template using the QRS complex segments;
calculate correlation coefficients between the QRS complex template and the QRS complex segment;
comparing the variability metric to a variability threshold and the correlation coefficients to a correlation threshold, and
designating the CA signals to include a predetermined level of PVC burden based on the comparing.
12. The system of claim 11 , wherein the one or more processors are configured to calculate the QRS scores comprises calculating at least one of a summation, area under a curve, or energy for the QRS complex segments.
13. The system of claim 11 , wherein the one or more processors are configured to create a morphology ensemble from the QRS complex segments; and compare the morphology ensemble to QRS morphologies for the corresponding QRS complex segments to obtain morphology correlation characteristics for the QRS complex segments, the designation based on the morphology correlation characteristics.
14. The system of claim 13 , wherein the morphology correlation characteristic represents an average correlation coefficient, and wherein the determining includes determining whether the average correlation coefficient is smaller than the predetermined threshold.
15. The system of claim 13 , wherein the morphology correlation characteristic represents a minimum correlation coefficient, and wherein the determining includes determining whether the minimum correlation coefficient for the QRS complex segments is smaller than a minimum threshold.
16. The system of claim 13 , wherein the morphology ensemble represents an ensemble average of the QRS morphologies for a desired number of the QRS complex segments; and the comparing includes comparing the QRS morphologies for the corresponding QRS complex segments to the ensemble average to obtain the morphology correlation characteristics.
17. The system of claim 11 , wherein the one or more processors are configured to compute the variability metric by calculating a covariance Cov_QRS SUM from the QRS scores, the covariance based on the QRS complex segments.
18. The system of claim 11 , wherein, when the variability metric does not satisfy the variability threshold, declaring the CA signals to include non-significant PVC burden which is less than the predetermined level of PVC burden.
19. The system of claim 1 , wherein, when the correlation coefficients do not satisfy the correlation threshold, declaring the CA signals do not have significant PVC burden.
20. The system of claim 1 , wherein, when the variability metric satisfied the variability threshold and the correlation coefficients satisfy the correlation threshold, declaring the CA signals to have significant PVC burden.
21. The system of claim 20 , wherein, when declaring the CA signals to have significant PVC burden, rejecting an original detection of AF episode, or reevaluating if the CA signals include AF by excluding identified PVC beats.