Direct discrimination and full-day disclosure of ventricular parasystole on new heart rate tachograms

The inverse problem for cardiac arrhythmias

A cardiac arrhythmia is an abnormality in the rate or rhythm of the heart beat. We study a type of arrhythmia called a premature ventricular complex (PVC), which is typically benign, but in rare cases can lead to more serious arrhythmias or heart failure. There are three known mechanisms for PVCs: reentry, an ectopic focus, and triggered activity. We develop minimal models for each mechanism and attempt the inverse problem of determining which model (and therefore which mechanism) best describes the beat dynamics observed in an ambulatory electrocardiogram. We demonstrate our approach on a patient who exhibits frequent PVCs and find that their PVC dynamics are best described by a model of triggered activity. Better identification of the PVC mechanism from wearable device data could improve risk stratification for the development of more serious arrhythmias.

Ectopic cycle length estimation from the quantified distribution patterns of ventricular bigeminy and trigeminy

Background

Ectopic cycle length (ECL) and the distribution patterns of ventricular bigeminy and trigeminy, expressed as their postextrasystolic intervals (PEIs) and interectopic intervals (IEIs), have been poorly pursued.

Objective

Based on modulation theory, we hypothesized that the PEIs of bigeminy and trigeminy determine their IEIs and ECL.

Methods

Ambulatory electrocardiograms of 1290 patients with ventricular premature complexes (≥3000/day) were studied. To quantify their distribution pattern on the PEI vs IEI curve (PIC), we introduced the following 2 ratios: PEI of trigeminy to PEI of bigeminy ratio (T/B-PEI) and IEI of trigeminy to IEI of bigeminy ratio (T/B-IEI). Distribution patterns were divided into 3 types by T/B-PEI: standard type (<0.90), intermediate type (between 0.90 and 1.20), and reverse type (>1.20). ECL was defined as the average of the bigeminy and trigeminy intervals in the standard type, and bigeminy intervals in the other 2 types.

Results

T/B-IEI disclosed significant linear relationship with T/B-PEI (P < .0001). ECLs were longest in the standard type (1905 ± 347 ms; n = 426), followed by the intermediate type (1520 ± 239 ms; n = 607) and reverse type (1317 ± 227 ms; n = 227) (P < .0001). Trigeminy PEI/ECL in the standard type (0.450 ± 0.074) was significantly shorter than that of the other 2 types (P < .0001).

Conclusion

We confirmed that T/B-PEI determines T/B-IEI and ECL by discriminating the 3 distribution patterns. Among them, trigeminy PEI/ECL decided the 2 types of modulation by the first sinus QRS, starting at the early delay phase or the later acceleration phase.

Strong modulation of ectopic focus as a mechanism of repetitive interpolated ventricular bigeminy with heart rate doubling

BACKGROUND

Repetitive interpolated ventricular bigeminy (RIVB) can introduce a doubling of the ventricular rate.

OBJECTIVE

To clarify the mechanism of RIVB, we hypothesized that it was introduced by a strong modulation of the ventricular automatic focus.

METHODS

RIVB, defined as more than 7 bigeminy events, was detected by instantaneous heart rate and bigeminy interval (BI) tachograms in 1450 successive patients with frequent ventricular premature contractions (≥3000 per day). Postextrasystolic interval bigeminy interval curves were plotted to determine the degree of modulation. Mean sinus cycle length bigeminy interval curves were plotted for selection. RIVB was simulated by using a computer-based parasystole model.

RESULTS

RIVB was observed in 7 patients (age 60 ± 16 years; 2 men and 5 women) with a heart rate of 58.2 ± 6.5 beats/min during a rest period both during the day and at night. The tachograms disclosed the onset of the RIVB with a doubled ventricular rate to 112.3 ± 8.5 beats/min. On the postextrasystolic interval bigeminy interval curves, compensatory bigeminy and interpolated bigeminy constituted overlapping regression lines with slopes close to 1.00 and RIVB was located in the lower left portion. RIVB lasting for up to 3 hours was quickly detected by mean sinus cycle length bigeminy interval curve. The PQ interval immediately after RIVB was prolonged in comparison with baseline (0.18 ± 0.02 to 0.21 ± 0.02 seconds; P < .001). The simulation was able to reproduce RIVB faithfully at a slow heart rate.

CONCLUSIONS

Our findings support the hypothesis that RIVB was introduced by strongly modulated ventricular pacemaker accelerated by an intervening normal QRS.

Variation in parasystolic cycle length

At the time of the first visit to our clinic, an electrocardiographic examination of a 73-year-old female patient revealed ventricular premature contractions (VPCs) with variable coupling intervals that were diagnosed as parasystole. Characteristically many of the parasystoles had no sinus contractions between two consecutive VPCs, which we referred to as pure parasystole. We first repeatedly examined variations in the length of the parasystolic cycles between January 6, 1997 and March 2, 2003 using electrocardiography. The time courses recorded over this period showed that the length of the parasystolic cycle did not remain constant, but varied irregularly within a relatively narrow range. We also recorded the length of the parasystolic cycles over 3 hours using Holter monitoring. The interectopic intervals plotted against mean sinus cycle length showed that the cycle length of pure parasystoles remained almost constant at about 1,300 ms over the 3 hours. We also examined the cycle length during exercise and found that it was slightly prolonged thereafter, while the sinus cycle length was clearly shortened after exercise. The average of six deep breathing tests showed that parasystolic cycle length did not significantly differ between deep inspiration and deep expiration, whereas the sinus cycle length during expiration was significantly longer than that during inspiration. These results indicate that the responses to both exercise and deep breathing obviously differed between the parasystolic and sinus cycle lengths.

Model of bidirectional modulated parasystole as a mechanism for cyclic bursts of ventricular premature contractions

Cyclic bursts of ventricular premature contractions (VPC) coming at minute-order intervals have been discerned by analyzing ambulatory ECG recordings, and their mechanism has not been clarified. The present study simulates this phenomenon by constructing a bidirectional modulated parasystole model. With Ts and Te as the intrinsic periods of the sinus and ectopic pacemakers, there are distinct and initial condition-dependent solutions in the model with Ts / Te values close to 1, 1/2, 1/4, etc. Typically, two distinct stable solutions are found existing together around Ts / Te = 1/2. We have verified theoretically the coexistence of different solutions and their dependence on the model parameters. The solution presented switches between those by a premature stimulus and those by fluctuations in the model parameters such as Ts. Patterns of RR intervals were generated by simulation with randomly fluctuating Ts. They included cyclic bursts of bigeminy of the "flat type" and the "dome type" reported by Takayanagi et al. (1999) and other transient types with Wenckebach or reverse Wenckebach rhythm of coupling intervals. This model provides a mathematical representation of the atrioventricular feedback mechanism and enables the modulated parasystole hypothesis to be applied to wider classes of VPCs.

Complex patterns of abnormal heartbeats

Individuals having frequent abnormal heartbeats interspersed with normal heartbeats may be at an increased risk of sudden cardiac death. However, mechanistic understanding of such cardiac arrhythmias is limited. We present a visual and qualitative method to display statistical properties of abnormal heartbeats. We introduce dynamical "heartprints" which reveal characteristic patterns in long clinical records encompassing approximately 10(5) heartbeats and may provide information about underlying mechanisms. We test if these dynamics can be reproduced by model simulations in which abnormal heartbeats are generated (i) randomly, (ii) at a fixed time interval following a preceding normal heartbeat, or (iii) by an independent oscillator that may or may not interact with the normal heartbeat. We compare the results of these three models and test their limitations to comprehensively simulate the statistical features of selected clinical records. This work introduces methods that can be used to test mathematical models of arrhythmogenesis and to develop a new understanding of underlying electrophysiologic mechanisms of cardiac arrhythmia.