RU2485880C2 - Method of predicting capture beat in patients with atrial fibrillation - Google Patents

Abstract

FIELD: medicine.

SUBSTANCE: invention relates to medicine, namely to cardiology. Indices of temporal and spectral HRV analysis are determined by analysis of 5-minute ECG registration at the background of arrhythmia. "Trees of solution" algorithm is built by means of software "Diagnostic system of capture beat prediction in case of atrial fibrillation based of heart rhythm variability analysis (DISC 3)". Prediction is carried out by comparison of time and frequency indices of patient's HRV with their threshold values. After that patients with arterial fibrillation were divided into two groups depending on arrhythmia outcome: with favourable and unfavourable prediction for capture beat.

EFFECT: method makes it possible to carry out early prediction of capture beat.

2 ex

Description

The invention relates to medicine, namely to cardiology. According to the literature, the factors determining the possibility of restoring sinus rhythm during atrial fibrillation include the duration of the arrhythmia, the size of the heart cavities, especially the left atrium, the presence of pathological conditions that contribute to rhythm disturbance (thyroid dysfunction, electrolyte imbalance, inflammatory processes in the body and etc.). None of these factors allows predicting the restoration of sinus rhythm with paroxysm of atrial fibrillation in the early stages of a particular patient, since all the methods used for examining patients (echocardiography, laboratory tests) require time.

There is evidence in the literature that in patients with a constant form of atrial fibrillation, heart rate variability is less than with a paroxysmal form of arrhythmia. A decrease in the variance of heart rhythm with an increase in the cavities of the heart was noted, the dependence of the standard deviation on the duration of arrhythmia was established (I. Kabanova // Statistical analysis of heart rhythm in patients with different clinical course of atrial fibrillation. Heart rate variability: theoretical aspects and practical application. International Symposium, Abstracts of reports. - Izhevsk, 1996, p. 35-36). M. van den Berg et al. noted the relationship between the variability of ventricular contractions and vagus nerve tone in patients with persistent atrial fibrillation (van-den-Berg MP, Haaksma J., Brouwer J. et al. Heart rate variability in patients with atrial fibrillation is related to vagal tone. Circulation. 1997 Aug 19; 96 (4): 1209-1216). But the possibility of using the criteria for heart rate variability (HRV) to predict the restoration of sinus rhythm with atrial fibrillation is not described in the literature.

The technical result of the proposed method consists in the possibility of early prediction of restoration of sinus rhythm in patients with paroxysm of atrial fibrillation by analyzing indicators of heart rate variability during arrhythmia using the DISK_3 computer program (Diagnostic system for predicting restoration of sinus rhythm in atrial fibrillation based on analysis of heart rhythm variability ". Certificate of state registration of the computer program No. 20100612460 dated 04/07/2010, authors - Ballod B.A., Muromkina A.V.).

In our proposed method, a 5-minute recording of the electrocardiogram of patients with atrial fibrillation was performed. With the help of the agricultural complex “Polyspectr-8 / EX” (Neurosoft, Ivanovo), an analysis of HRV was performed against the background of arrhythmia. The parameters of spectral and temporal analysis of HRV were estimated: the total spectrum power (TP), the ratio of high frequency (HF), low frequency (LF) and very low frequency (VLF) waves, the average duration of the interval between cardiocycles (RRNN), the standard deviation is the square root from the range of RR intervals (SDNN), as well as the ratio of standard deviation to the average duration of the interval between cardiocycles (SDNN / RRNN). The obtained indicators were entered in the dialogue form of the program DISK_3. This program performs the task of classifying objects into groups by constructing "Decision Trees" in the form of Classification And Regression Tree (C&RT) ("Classification and Regression Tree"). This classifier is based on the analysis of HRV indices in a training sample of 110 patients with various forms of atrial fibrillation. Among the temporal parameters of HRV, the highest rank of significance for predicting the outcome of arrhythmia was obtained for SDNN and SDNN / RRNN, and among the frequency parameters for TP, HF, LF, and VLF. C&RT consists of leaves pointing to the class and nodes containing classification rules. It can be used to classify objects that are not included in the training set. In the C&RT algorithm, each node of the decision tree has two children: right is the part in which the rule is executed, and left is the part in which the rule is not executed. To determine the factor by which the partition will occur, the entropy indicator is used - a measure of uncertainty. The factor is selected that, when partitioned, eliminates more uncertainty. Entropy is zero if the node contains objects belonging to the same class. Building a decision tree using the C&RT algorithm (exhaustive search) is implemented by the corresponding module of the statistical analysis package STATISTICA. This program allows you to assign the patient to one of two groups: with a favorable or unfavorable prognosis for the restoration of sinus rhythm with a probability of about 90% (the number of errors is less than 10% of the number of diagnosed). Based on a preliminary analysis of HRV and a decision of the intellectual system, the doctor makes a conclusion about the likelihood of restoration of sinus rhythm in a particular patient with atrial fibrillation.

Using this method of predicting rhythm recovery allows you to reduce the time spent on the initial examination of patients with atrial fibrillation and to identify a group of patients with a high probability of restoration of sinus rhythm. The proposed method is quite simple, safe for patients, does not require large expenditures of time and complex equipment, can be used both in inpatient and outpatient settings.

Clinical example No. 1.

Patient A., 58 years old. The medical history No. 031331.

Was admitted to the hospital with paroxysm of atrial fibrillation lasting 5 hours, HRV analysis was performed, the data were entered in the dialogue form of the DISK_3 program. The result of the forecast: "restoration of rhythm."

SPECTRAL ANALYSIS TEMPORARY ANALYSIS FORECAST Name of patient TR Vlf Lf Hf Rrnn Sdnn SDNN / RRNN branch 1 branch 2 decision Ale ... 13468 1796 4295 7377 499 126 0.25 AT RESTORED

Treatment: taking into account the tachysystole that has taken place, metoprolol is prescribed to control the number of heart contractions. Against this background, after 3 hours, sinus rhythm was restored. Match treatment results and prognosis.

Clinical example No. 2.

Patient N., 58 years old. The history of the disease №031306.

He was admitted to the hospital with the first detected episode of atrial fibrillation of unknown age. The analysis of HRV, the data are entered in the dialog form of the program DISK_3. The result of the forecast: "arrhythmia".

SPECTRAL ANALYSIS TEMPORARY ANALYSIS FORECAST Name of patient TR Vlf Lf Hf Rrnn Sdnn SDNN / RRNN branch 1 branch 2 decision New ... 12422 685 5554 6183 544 125 0.23 BUT ARRHYTHMIA

Treatment: taking into account the unknown prescription of arrhythmia, metoprolol was prescribed to control the number of heart contractions. According to the results of a subsequent examination by echocardiography revealed dilatation of the left atrium, it was decided to abstain from the appointment of antiarrhythmic drugs. The outcome of the disease is the preservation of a constant form of arrhythmia. Match treatment results and prognosis.

Claims (1)

A method for predicting the restoration of sinus rhythm in patients with atrial fibrillation, including the assessment of heart rate variability indicators, characterized in that the analysis of heart rate variability indicators against the background of atrial fibrillation using the computer program "Diagnostic system for predicting rhythm recovery during atrial fibrillation based on analysis of heart rate variability (DISK_3) ”, which using the“ decision trees ”algorithm allows to separate patients with fibrillation atrial fibrillation at the end of two groups: those with a favorable or unfavorable prognosis for the restoration of sinus rhythm.