Annotation: a normal heart beat as N, an interpolated ventricular premature contractions as V, and heart beat using atrioventricular node slow pathway as N′.
Introduction
A Lorenz plot is a scatter plot on which each R-R interval is plotted against the next R-R interval. To provide a global view of long-time electrocardiogram information, the Lorenz plot has been regarded as a simple visualization tool to analyze cardiac arrhythmias. A three-dimensional Lorenz plot (3DLPs) is constructed with three successive R-R intervals as the X-axis, Y-axis, and Z-axis in a Cartesian coordinate system for each record. The stereographic projection of 3DLPs was along with the space diagonal. The R-R interval series are described as a time series vector [x 1, x 2,.., xn ]T. The 3DLP is constructed by a plot of xi -1 on the X-axis, xi on the Y-axis, and xi +1 on the Z-axis in the three-dimensional Cartesian coordinate system of (xi -1, xi , xi +1). The algorithm for 3DLPs is used to calculate the attractor feature of the three interphase R-R intervals of four beats. Cluster extraction is done by automatic segmentation of the 3DLP based on homogeneously statistical properties of each coordinate element. So the morphological characteristics of the 3DLP might be different for different arrhythmias. What is more, the clusters’ space contours are more distinct for less overlap with each other. Combined with the trajectory tracking technique, the 3DLP has the significant advantage of extraction of attractors and recognition of arrhythmias. An interpolated premature ventricular contraction (PVC) is sandwiched between two consecutive sinus beats without interfering with the sinus rhythm. This type of arrhythmia usually occurs when the sinus rate is slow and the PVC is early. The P-R interval of the sinus complex following the PVC is always prolonged because of concealed retrograde conduction of the ectopic ventricular impulse, which renders the atrioventricular junction partially refractory to the closely following sinus impulse. In our study, we present the specific 3DLP pattern of a patient with an atrioventricular node double path caused by interpolated ventricular premature contractions.
Case Presentation
Clinical Data
A 65-year-old man who had experienced acute extensive anterior myocardial infarction underwent 24-hour Holter monitoring (DM Software, Stateline, NV, USA). Cluster extraction was done by automatic segmentation of 3DLPs (DM Software) based on homogeneously statistical properties of each coordinate element. All records had been verified carefully by experienced cardiologists. This study complied with the Declaration of Helsinki and was approved by the local ethics committees of Chizhou Second People’s Hospital, with the patient giving written informed consent.
First, there are six distributions in the standard surface of observations (XOY) by 3DLPs based on the recorded data: (1) a PVC precluster, (2) an interpolated premature beat cluster, (3) a common PVC cluster, (4) and interpolated premature beat postcluster, (5) a common single PVC postcluster, and (6) a normal heart beat cluster (Figure 1A). Second, from the top overlooking surface of isovelocity (XYZ) by 3DLPs, we can see that two independent subclusters separated from the interpolated premature beat precluster (Figure 1B). Third, from the surface of observations (OYZ) by 3DLPs, two independent subclusters separated from the interpolated premature beat cluster. One subcluster consists of interpolated ventricular premature contractions conducted by the normal atrioventricular node pathway. The other one consists of interpolated ventricular premature contractions conducted by the slow atrioventricular node pathway (Figure 1C). Fourth, from the surface of observations (XYO) by 3DLPs, we also found two independent subclusters separated from the interpolated premature beat postcluster (Figure 1D). Fifth, combined with the trajectory tracking function, we found only two P-R intervals with significant differences after the interpolated PVC. The P-R interval of R4 was 210 ms, and the P-R interval of R8 was 420 ms. This leap phenomenon suggests the atrioventricular node double conduction path. It is considered that R8 is conducted by the slow pathway of the atrioventricular junction (Figure 2). Finally, we conducted this observational study of 3DLPs for a patient without atrioventricular node double path caused by interpolated ventricular premature contractions recently. For the patient without atrioventricular node double path caused by interpolated ventricular premature contractions, there are also six distributions in the standard surface of observations (XOY) by 3DLPs based on the recorded data: (1) a PVC precluster, (2) an interpolated premature beat cluster, (3) a common PVC cluster, (4) an interpolated premature beat postcluster, (5) a common single PVC postcluster, and (6) a normal heart beat cluster (Figure 3A). From the surface of observations (OYZ) by 3DLPs, there is only one cluster from the interpolated premature beat (Figure 3B). From the surface of observations (XYO) by 3DLPs, we detected only one cluster composed of the interpolated premature beat postcluster (Figure 3C). From these two points, we can make a distinct decision.
(A) There are Six Distributions in the Standard Surface of Observations (XOY) by Three-dimensional (3D) Lorenz Scatter Plots: (1) Premature Ventricular Contraction Precluster (consists of beats before premature ventricular contractions), (2) Interpolated Premature Beat Cluster (consists of interpolated premature beats), (3) Common Premature Ventricular Contraction Cluster (consists of common premature ventricular contractions), (4) Interpolated Premature Beat Postcluster (consists of beats after interpolated premature beats), (5) Common Single Premature Ventricular Contraction Postcluster (consists of beats after common single premature ventricular contractions), and (6) Normal Heart Beat Cluster (consists of normal heart beats). (B) From the Top Overlooking Surface of Isovelocity (XYZ) by 3D Lorenz Scatter Plots, We See Two Independent Subclusters Separated From The Interpolated Premature Beat Precluster. The Green Cluster Corresponds to the Atrioventricular Node Normal Pathway (NNVN; N corresponds to a normal heart beat and V corresponds to an interpolated ventricular premature contraction), While the Red Cluster Corresponds to the Atrioventricular Node Slow Pathway (NNVN′; N′ corresponds to a heart beat using the atrioventricular node slow pathway). (C) From the Surface of Observations (OYZ) by 3D Lorenz Scatter Plots, Two Independent Subclusters Separated from Interpolated Premature Beat Cluster. The NVNN Subcluster Consists of Interpolated Ventricular Premature Contractions Conducted by the Normal Atrioventricular Node Pathway. The NVN′N Subcluster Consists of Interpolated Ventricular Premature Contractions Conducted by the Slow Atrioventricular Node Pathway. (D) From the Surface of Observations (XYO) by 3D Lorenz Scatter Plots, We Also Found Two Independent Subclusters Separated From the Interpolated Premature Beat Postcluster. The VNNN Subcluster Consists of an Interpolated Premature Beat Postcluster That is Conducted by the Normal Atrioventricular Node Pathway. The VN′NN Subcluster Consists of an Interpolated Premature Beat Postcluster That is Conducted by the Slow Atrioventricular Node Pathway.
With Use of the Trajectory Tracking Function, the P-R Interval of R4 was 210 ms and the P-R Interval of R8 was 420 ms.
It is considered that R8 is conducted by the slow pathway of the atrioventricular junction.
(A) For the Patient Without Atrioventricular Node Double Path Caused by Interpolated Ventricular Premature Contractions, There Are Also Six Distributions in the Standard Surface of Observations (XOY) by Three-dimensional (3D) Lorenz Scatter Plots Based on the Recorded Data: (1) Premature Ventricular Contraction Precluster, (2) Interpolated Premature Beat Cluster, (3) Common Premature Ventricular Contraction Cluster, (4) Interpolated Premature Beat Postcluster, (5) Common Single Premature Ventricular Contraction Postcluster, and (6) Normal Heart Beat Cluster. (B) From the Surface of Observations (OYZ) by 3D Lorenz Scatter Plots, There is Only One Cluster from the Interpolated Premature Beat. (C) From the Surface of Observations (XYO) by 3D Lorenz Scatter Plots, We Could Detect Only One Cluster Composed of the Interpolated Premature Beat Postcluster. N, Normal Heart Beat; V, Interpolated Ventricular Premature Contraction.
Discussion
Interpolated ventricular premature contraction is not an uncommon event in dynamic electrocardiogram analysis and may cause some special arrhythmias [1–3]. However, because of the massive amount of data recorded by dynamic electrocardiogram, the effects of atrioventricular node double path caused by interpolated ventricular premature contraction are widely ignored. Our present study was aimed at filling this gap. The P-R interval of the sinus rhythm following the interpolated ventricular contraction is nearly always prolonged because of concealed conduction of the ectopic ventricular impulse, which affects the atrioventricular junction relative refractory period. There was good correlation between the P-R interval due to the concealed conduction and the coupling interval of the interpolated ventricular conduction, sinus P-P cycle length, and following R-P interval. Under the circumstance of consecutive sinus rhythm, the same coupling interval of the interpolated ventricular conduction and the same R-P interval following the interpolated ventricular contraction, there are only two kinds of P-R interval of the sinus rhythm following the interpolated ventricular contraction (that is, 210 ms and 420 ms) from the patient’s 24-hour Holter monitoring. Combined with the trajectory tracking strategy of 3DLPs, we found the character of the corresponding electrocardiogram feature did not support concealed conduction. On the contrary, it is the characteristic manifestation of the atrioventricular node double conduction path. So we conclude the presence of atrioventricular node double conduction paths. On the basis of the separation characteristics of the associated cluster, 3DLPs are useful for revealing patients with atrioventricular node double path caused by interpolated ventricular premature contractions. We found two independent subclusters separated from the interpolated premature beat precluster, the interpolated premature beat cluster, and the interpolated premature beat postcluster respectively. Cluster extraction was done by automatic segmentation of 3DLPs based on homogeneously statistical properties of each coordinate element. So the morphological characteristics of 3DLPs might be different for different arrhythmias. Previously, we conducted an observational study of 3DLPs for a patient without atrioventricular node double path caused by interpolated ventricular premature contractions. From the surface of observations by 3DLPs, there is only one cluster from the interpolated premature beat and the interpolated premature beat postcluster. From these two points, we can make a distinct decision.
Dual atrioventricular nodal pathways (DAVNPs) are a common electrophysiological phenomenon. The mechanisms of atrioventricular nodal reentrant tachycardia, reciprocal rhythm, dual atrioventricular nodal non-reentrant tachycardia, and prolonged P-R interval syndrome are associated with a DAVNP. Three-dimensional Lorenz scatter plots make it possible for patients with DAVNP to receive a diagnosis early and be treated early. Three-dimensional Lorenz scatter plots have the significant advantage of extraction and recognition of clusters for significantly fewer overlaps than two-dimensional Lorenz scatter plots [4–6]. The algorithm for 3DLPs is to calculate the three interphase R-R intervals of four beats, with the advantage of free rotation. Combined with the trajectory tracking technique, we can find the atrioventricular node double conduction path caused by interpolated PVC. Because of this characteristic morphological image, it is easier to identify with reciprocal ventricular beats or paired ventricular premature beats. Three-dimensional Lorenz R-R interval scatter plots can be regarded as a combination of conventional two-dimensional Lorenz R-R interval scatter plots and difference scatterplots. Moreover, 3DLPs also have the advantage of free rotation. It is concluded that in the context of both 24-hour Holter monitoring and the trajectory tracking technique, the 3DLP method has the potential to significantly increase the accuracy of arrhythmia detection and differentiation. Our results provide a sound framework for a meaningful interpretation of the atrioventricular node double path caused by interpolated ventricular premature contractions. Up to now, the application time of 3DLPs in a dynamic electrocardiogram has not been long. Therefore, further methods to extract clusters from three-3DLPs are required to confirm an effective clinical application.
Conclusion
Through the characteristic morphological manifestations of 3DLPs, we can give a preliminary assessment of the existence of the atrioventricular node double path. Combined with the trajectory tracking technique, we can find the atrioventricular node double conduction path caused by interpolated PVC.
Limitations
First, the physician suggested the patient undergo electrophysiological studies or a transesophageal atrial pacing study; however, the patient did not follow the physician’s advice for personal reasons.
Second, we conducted this observational study of 3DLPs for detecting patients with atrioventricular node double path caused by interpolated ventricular premature contractions recently. With the development of clinical research into 3DLPs, the sensitivity, specificity, and positive and negative predictive accuracy of this method should be thoroughly evaluated.