RU2728869C1 - Computerized method for non-invasive detection of carbohydrate metabolism disorders by electrocardiogram - Google Patents

Abstract

FIELD: medicine.SUBSTANCE: group of inventions relates to medicine, specifically to computerized methods for non-invasive detection of carbohydrate metabolism disorders (CMD) by cardiac electrocardiogram (ECG), population screening for detecting individuals with CMD signs by their ECG, as well as monitoring the level of the patient's CMD by ECG. Heart is considered as an open non-linear self-oscillation system characterized by Fermi-Pasta-Ulam (FPU) self-return, and the ECG is a representation of a FPU automatic return. ECG is removed. Fourier transform of the recorded ECG is performed to obtain an amplitude Fourier spectrum. Fourier amplitude spectrum is discretised to obtain discrete form of initial form of Fourier amplitude spectrum (DECG). DECG is analysed by computer processing to detect the presence/absence of the cardiac CMD by establishing the similarity/difference with the DECG of the patients with the CMD from an annotated sample of ECG and DECG of the healthy patients from the annotated ECG sample. Presence of similarity of analysed DECG with DECG of patients with CMD and absence of similarity with DECG of healthy patients is a sign of CMD. Screening method further includes setting sensitivity targets and specificity for determining the minimum required number of ECG of one person being tested and a threshold value of the number of ECG with signs of an CMD for each age group of observation, exceeding which the presence of an CMD is recognized in the patients of the group. Population is screened by ECG data extraction and processing. Results are statistically analysed by determining from the minimum required ECG number of one analysed ECG count with the CMD feature and comparing it with the preset threshold value. Monitoring method additionally involves accumulation in the database of the sequence of ECG recorded marked with patient identifiers. Degree of difference of the analysed DECG from the DECG of healthy people is calculated and used as an integral parameter of the CMD. Time variation of the level of the patient's CMD of the observed patient is evaluated by measuring the time-varying integral index. Level of the CMD is predicted on the basis of the "observation history", which is the recorded changes in time of previous values of the integral index.EFFECT: higher accuracy of assessing the patient's cardiac state for more qualitative detection of CMD by ECG, as well as reducing the time for diagnosing, screening the population and monitoring the level of the CMD.7 cl, 2 tbl, 4 ex, 7 dwg

Description

Technology area

The present invention relates to the field of medicine, and more precisely to a computerized method for detecting disorders of carbohydrate metabolism by an electrocardiogram (ECG) of the heart, which is considered as an open nonlinear self-oscillatory system characterized by Fermi-Pasta-Ulam (FPU) auto-return, while the ECG is a representation auto return FPU.

The method can be used to diagnose disorders of carbohydrate metabolism and further monitor the patient to diagnose the presence / absence of diabetes mellitus. The method can be used in various settings, including clinical hospitals, paramedic centers, as well as to monitor the condition of patients outside the hospital on outpatient treatment.

Prior art

Type 2 diabetes mellitus (T2DM) is a chronic endocrine disease that develops as a result of insulin resistance and dysfunction of the beta cells of the pancreas.

Type II diabetes mellitus is based on a violation of carbohydrate metabolism due to an increase in cell resistance to insulin (insulin resistance). The ability of tissues to accept and utilize glucose decreases, a state of hyperglycemia develops - an increased level of plasma sugar.

The difficulty of detecting non-insulin dependent diabetes mellitus is explained by the absence of pronounced symptoms at the initial stages of the disease. In this regard, people at risk and all people over 40 years of age are recommended to screen for plasma sugar levels. Laboratory diagnostics is the most informative, it allows detecting not only the early stage of diabetes, but also the state of pre-diabetes.

Improving the quality of medical care for patients with various diseases at an acceptable cost of developing the health care system is an extremely urgent task. The most important direction of this work is the development of various methods of telemedicine for diagnostics, screening, monitoring and accompanying patients, regardless of their territorial location. One of these diseases is diabetes mellitus.

The known method and device for non-invasive diagnosis of patients (see, for example, RU 2487655 C2, published on July 20, 2013). The method consists in measuring the patient's biophysical parameters at the diagnostic point, transferring them to a unified system for processing and accumulating information, registering the patient's measurement results in it, taking into account his identification data, analyzing and establishing a diagnosis, converting it into a feedback signal arriving at the diagnostic point. In this case, at the diagnostic point for each patient, a fixed list of biophysical parameters is measured and an informative vector of the patient's parameters obtained as a result of the current diagnostic examination is formed.

Diagnostic examinations are carried out periodically, and according to the difference in the informative vectors of the patient's parameters as a result of the current and previous diagnostic examinations, an informative vector of deviation of parameters is determined to determine the degree of change in the patient's condition, in accordance with which, by the patient's identification data and an informative vector reflecting the history of the disease and hereditary factors, make a diagnosis. In case of insufficient information for unambiguous determination of the diagnosis at the diagnostic point, additional measurements of biophysical parameters that are not included in the fixed list are carried out, an additional informative vector of the patient's parameters is formed and a diagnosis is made taking it into account. Registration of the results of measurements of biophysical parameters is carried out in a single register for all patients included in the medical support system.

The system contains m diagnostic points, which accommodate diagnostic equipment for non-invasive measurements, a communication channel with the information processing and accumulation unit, a diagnostic evaluation unit, a monitoring, control and information display unit, a unit for automatic determination of methods and means of treatment. All diagnostic equipment is connected to the corresponding inputs of the output adapter at each diagnostic point, and its output is connected by a duplex communication channel to the input adapter of the system, which is connected by the first output to the first input of the diagnostic evaluation unit and to the first input of the information processing and accumulation unit, the second output to the first input of the block for additional monitoring of the patient's condition. Diagnostic equipment for additional measurement of biophysical parameters is connected via a communication channel to the corresponding input of the additional monitoring unit for patients. The use of the invention makes it possible to increase the objectivity of diagnosing patients by automating the decision-making processes.

The disadvantages of this device include the need to visit a diagnostic center, repeated diagnostic examinations to detect deviations in parameters to determine the degree of change in the patient's condition, according to which, taking into account hereditary factors, a diagnosis is made.

A device for cardiographic monitoring of patients' condition is considered as the closest technical solution (see, for example, RU 2615721 C2, published on 04/07/2017). The device contains a monitor, interface, ECG electrodes for removing electrocardiographic signals from the patient's body, connected to the input of the primary signal processing unit, the other input of this unit is connected to the output of the time sampling unit, and the output of the primary signal processing unit is connected to the channel switching unit. The outputs of the channel switching unit are connected to the discrete Fourier transform unit, the output of which contains the values of the amplitude, frequency and phase of the harmonics of the signal under study, and to the patient data input unit. Harmonics are processed in the cardiogram fixer, which stores and outputs the required amount of time at the output of the harmonic amplitude of the signal under study. The amplitudes of the harmonics are fed to the cardiogram image determinant, which compares the resulting image from the ECG electrode, taking into account the confidence intervals and a certain degree of reliability, with the images from the base of the cardiogram images.

The output of the determinant is connected to the input of the block for fixing states and analyzing their dynamics, where, according to the data of the images of the cardiograms from all ECG electrodes, a diagnosis of the patient's disease is formed by comparing the set of images of the cardiograms from the ECG electrodes with the set characterizing the diagnosis of the disease from the base of cardiological diagnoses, taking into account the confidence intervals and a certain degree of reliability. In the same block, the degree of reliability of the diagnosis, the dynamics of the diagnosis depending on the previous study of the patient, and the time of determining the diagnosis are determined. The data are displayed on the monitor, transferred to the interface for storage and research on other technical means and to the patient data input unit, where they are stored in the corresponding patient archives.

The specified device has a complex design and the criteria used to assess the disorders of carbohydrate metabolism by the cardiac state of the patient are not disclosed.

Statement of the essence of the invention

IN The basis of the present invention is the task of creating a computerized method for non-invasive detection of disorders of carbohydrate metabolism on the electrocardiogram (ECG) of the heart, which is considered as an open nonlinear self-oscillatory system characterized by Fermi-Pasta-Ulam (FPU) auto-return, while the ECG is a representation of the auto-return of the FPU. The present invention is also based on the task of creating a computerized method for screening the population to identify individuals with signs of carbohydrate metabolism disorders, and a computerized method for monitoring the level of carbohydrate metabolism disorders in a patient.

The technical result to which the present invention is directed is to improve the accuracy of assessing the patient's cardiological state for better detection of carbohydrate metabolism disorders by electrocardiogram, as well as to reduce the cost and shorten the time for diagnostics, screening of the population and monitoring the level of carbohydrate metabolism disorders in the patient.

The specified technical result is achieved by creating a computerized method for non-invasive detection of disorders of carbohydrate metabolism on the electrocardiogram (ECG) of the heart, which is considered as an open nonlinear self-oscillatory system, characterized by Fermi-Pasta-Ulam (FPU) auto-return, while the ECG is a representation of the auto-return of the FPU, this method contains stages at which:

- carry out the removal of at least one electrocardiogram of at least the first lead;

- carry out the Fourier transform of the specified at least one ECG to obtain the amplitude Fourier spectrum for the specified at least one ECG;

- produce discretization of the Fourier amplitude spectrum of the specified at least one ECG to obtain a discrete form of the original form of the Fourier amplitude spectrum (DECG) of the specified at least one ECG;

- use the obtained DECG for computer processing, in which

- analyze DECG to detect the presence / absence of a cardiac sign of carbohydrate metabolism disorder (MCI) by establishing similarity / difference with DECG of patients with MCI from an annotated sample of ECG and DECG of healthy patients from an annotated sample of ECG, while the presence of similarity of the analyzed DECG with DECG of patients with MCI and the lack of similarity with DECG in healthy patients is a sign of impaired carbohydrate metabolism.

Preferably, the sampling step is performed by constructing a Fourier envelope of the transformed at least one ECG in natural / averaged form.

Preferably, the sampling step is performed by constructing a harmonic peak envelope for the Fourier transformed at least one ECG.

The specified technical result is also achieved by creating a computerized screening method for the population to identify persons with signs of carbohydrate metabolism disorders by their electrocardiogram (ECG) of the heart, which is considered as an open nonlinear self-oscillating system, characterized by Fermi-Pasta-Ulam (FPU) auto-return, while the ECG is a representation of the automatic return of the FPU, the specified method contains the stages at which:

- set targets for the sensitivity and specificity of screening;

- carry out preparation for screening, for which it is determined according to the table, taking into account the set target indicators of sensitivity and specificity, the minimum required number of electrocardiograms (ECG) of one subject and the threshold value of the number of ECGs with signs of SME for each age group of observation, upon exceeding which the presence of ICR in patients of the group, and

- conduct screening of the population to identify persons with signs of impaired carbohydrate metabolism, for which

- the minimum required number of electrocardiograms (ECG) of at least the first derivation for one subject is removed, determined according to the table;

- carry out the Fourier transform of each captured ECG from a certain number of ECGs to obtain an amplitude Fourier spectrum for each captured ECG,

- discretization of the Fourier amplitude spectrum of the specified each captured ECG is performed to obtain a discrete view of the initial form of the Fourier amplitude spectrum (DECG) of the specified each captured ECG,

- use DECG for computer processing, in which

- analyze DECG to detect the presence / absence of a cardiac sign of carbohydrate metabolism disorder (MCI) by establishing similarity / difference with DECG of patients with MCI from an annotated sample of ECG and DECG of healthy patients from an annotated sample of ECG, while the presence of similarity of the analyzed DECG with DECG of patients with MCI and lack of similarity with the DECG of healthy patients is a sign of a violation of carbohydrate metabolism,

- a statistical analysis of the screening results is carried out, for which, from the minimum required number of ECGs of one subject, the number of ECGs with an ICG sign is determined and compared with a predetermined threshold value, by which the presence of a carbohydrate metabolism disorder in the subject is judged.

It is preferable that in the specified method the severity of the carbohydrate metabolism disorder is additionally assessed by how many ECGs with the ICG sign of one subject exceeds the threshold value, which results in a preliminary diagnosis of diabetes mellitus.

The specified technical result is also achieved by creating a computerized method for monitoring the level of carbohydrate metabolism disorders in a patient according to the electrocardiogram (ECG) of the heart, which is considered as an open nonlinear self-oscillatory system characterized by Fermi-Pasta-Ulam (FPU) auto-return, while the ECG is a representation of the auto-return of the FPU, the specified method contains stages at which:

- for each patient from the observed set of patients, the sequence of the collected set of his electrocardiograms (ECG) of at least the first lead marked with the patient identifiers is accumulated in the database, the specified sequence is obtained by adding each new recorded ECG of the patient marked with the indicated identifiers to its previously accumulated sequences of multiple ECGs throughout the observation period;

- carry out the Fourier transform of each of the captured set of ECGs to obtain the amplitude Fourier spectrum for each of the captured set of ECGs,

- discretization of the amplitude Fourier spectrum of each of the captured set of ECGs is performed to obtain a discrete form of the initial form of the amplitude Fourier spectrum (DECG) of each of the captured set of ECGs,

- use the obtained DECG for computer processing, in which

- analyze DECG to detect the presence / absence of a cardiac sign of carbohydrate metabolism disorder (MCI) by establishing similarity / difference with DECG of patients with MCI from an annotated sample of ECG and DECG of healthy patients from an annotated sample of ECG, while the presence of similarity of the analyzed DECG with DECG of patients with MCI and lack of similarity with the DECG of healthy patients is a sign of a violation of carbohydrate metabolism,

- calculate the degree of difference between the analyzed DECG and the DECG of healthy people and use it as an integral indicator of disorders of carbohydrate metabolism,

- the time change in the level of carbohydrate metabolism disturbance in the observed patient is estimated by the time change in the integral indicator of the carbohydrate metabolism disturbance

- prediction of the level of carbohydrate metabolism disorder of the observed patient is carried out on the basis of the "observation history", which is the registered changes in time of the previous values of the integral indicator of the carbohydrate metabolism disorder.

The specified technical result is also achieved by creating a computerized method for monitoring the level of carbohydrate metabolism disorders in a patient according to the electrocardiogram (ECG) of the heart, which is considered as an open nonlinear self-oscillatory system characterized by Fermi-Pasta-Ulam (FPU) auto-return, while the ECG is a representation of the auto-return of the FPU, the specified method contains stages at which:

- for each patient from the observed set of patients, the sequence of the collected set of his electrocardiograms (ECG) of at least the first lead marked with the patient identifiers is accumulated in the database, the specified sequence is obtained by adding each new recorded ECG of the patient marked with the indicated identifiers to its previously accumulated sequences of multiple ECGs throughout the observation period;

- carry out the Fourier transform of each of the captured set of ECGs to obtain the amplitude Fourier spectrum for each of the captured set of ECGs,

- discretization of the amplitude Fourier spectrum of each of the captured set of ECGs is performed to obtain a discrete form of the initial form of the amplitude Fourier spectrum (DECG) of each of the captured set of ECGs,

- use the obtained DECG for computer processing, in which

- analyze DECG to detect the presence / absence of a cardiac sign of carbohydrate metabolism disorder (MCI) by establishing similarity / difference with DECG of patients with MCI from an annotated sample of ECG and DECG of healthy patients from an annotated sample of ECG, while the presence of similarity of the analyzed DECG with DECG of patients with MCI and lack of similarity with DECG in healthy patients is a sign of impaired carbohydrate metabolism

- the time change in the level of carbohydrate metabolism disturbance in the observed patient is assessed by the frequency of occurrence of cardiograms with a sign of CHD

- prediction of the level of carbohydrate metabolism disturbance of the observed patient is carried out on the basis of the "observation history", which is the recorded changes in time of the previous values of the frequency of occurrence of cardiograms with a sign of ICR.

The proposed technical solution provides a reliable method for non-invasive detection of disorders of carbohydrate metabolism by electrocardiogram.

The proposed invention allows the patient to independently carry out preliminary diagnostics of the condition and quickly and unambiguously register violations of carbohydrate metabolism by electrocardiogram.

Brief Description of Drawings

The invention is further illustrated by the description of preferred embodiments thereof with reference to the accompanying drawings, in which:

FIG. 1 depicts the Fourier spectrum of an electrocardiogram (ECG) of a healthy patient, where the X-axis represents the frequency, the Y-axis represents the amplitude of the harmonics;

FIG. 2 depicts the Fourier spectrum of an electrocardiogram (ECG) with a sign of impaired carbohydrate metabolism;

FIG. 3a depicts DECG of healthy patients with a reliably known absence of signs of impaired carbohydrate metabolism;

FIG. 3c depicts DECG of inpatients with signs of impaired carbohydrate metabolism;

FIG. 4a depicts the dynamics of the ratio of ECG with and without a detected sign of CMD (patient with type 2 diabetes, 73 years, experience with type 2 diabetes),

FIG. 4c depicts the total proportion of ECGs with and without a symptom of CMD for the entire observation period (patient type 2, 73 years, experience type 2–25 years);

FIG. 5a depicts the dynamics of the ratio of ECG with and without a detected symptom of CMD (patient 73 years old, no T2DM);

FIG. 5c depicts the total ECG fraction with and without signs of CMD for the entire observation period (patient 73 years old, no T2DM);

FIG. 6a depicts the dynamics of the ratio of ECG with and without a detected sign of CMD (patient, 44 years old, no T2DM);

FIG. 6c depicts the total proportion of ECG with and without signs of CMD for the entire observation period (patient, 44 years old, no T2DM);

FIG. 7a depicts the dynamics of the ratio of ECG with and without a detected sign of CMD (patient 34 years old, no T2DM);

FIG. 7c depicts the total proportion of ECG with and without signs of CMD for the entire observation period (patient 34 years old, no T2DM).

Theoretical justification of the proposed method

In 1954, scientists Fermi, Pasta and Ulam made an attempt to numerically show the inevitability of the transition of a nonlinear system to a state with equal energy distribution from a state of monomode excitation.

The authors of the present invention proposed and mathematically described a version of the Fermi – Pasta – Ulam return model (FPU auto return) and put forward a hypothesis about an adequate description of the electrical dynamics of the heart within the discovered phenomenon of FPU auto return. Autoreturn is a pattern of the state of a nonlinear system repeating in time. As a result of computer study of the model, the dynamics of decisions of automatic return of the FPU was obtained, corresponding to the electrical dynamics of a normally functioning heart in the form of an electrocardiogram (ECG) (see A.V.Schmid, A.A. Berezin, M.A. Novopashin "Fermi-Pasta-Ulam auto-return in the description of the electrical activity of the heart ”(Fermi-Pasta-Ulam auto recurrence in the description of the electrical activity of the heart), Journal Medical Hypothesis, 101 (2017) pp. 17-22).

It is known that the heart is an open nonlinear self-oscillating system, the model of which, subject to a number of restrictions, can be a system of two coupled van der Pol equations.

A mass study using BIG DATA technologies of the first lead ECG from the Internet cardiograph CardioQwark, collected with constant monitoring of a group of patients for many months, revealed a number of patterns that do not fit into the specified model and are typical for both healthy people and patients with certain cardiopathologies.

In order to test the hypothesis, more than 20,000 ECGs of both healthy people and people with disorders of carbohydrate metabolism were examined.

As a result of processing ECG arrays of healthy patients, the characteristic type of Fourier ECG spectra in healthy people was determined.

To simulate the ECG, it was taken into account that the functioning of the heart occurs in a self-oscillating mode, which implies the presence of a similar principle of dynamics in the structure of the FPU return used to simulate cardiac activity.

As a result of these studies, it was found that the dynamics of the electrical activity of a normally functioning heart can be interpreted by the phenomenon of FPU auto-return, and disturbance of carbohydrate metabolism is characterized by the appearance of noise when energy flows between low-frequency and high-frequency harmonics of FPU auto return.

As shown by mathematical modeling, the electrical activity of the heart is the first formulated by the authors of the automatic return of the FPU, which is of an individual nature and contains in its structure a picture of the physiological state of the heart and pathological processes in the body, the patterns of which periodically appear in the Fourier spectra of automatic returns.

Auto returns of Fourier spectra of ECG of healthy people contain both a low-frequency part in the range of 1–5 Hz, corresponding to the canonical return of the FPU, and a high-frequency part in the range of 20–35 Hz, corresponding to a high-frequency return of the FPU. Thus, the ECG is a biological example of the complete auto return of FPU.

Based on the simulation results, it was suggested that in the set of frequency patterns obtained from different cardiograms of one patient or several patients, there is no chaotic distribution, and therefore it is possible to distinguish several groups of frequency patterns at some degree of proximity.

This means that clustering algorithms can be applied to a variety of ECG frequency patterns and stable frequency patterns of the states of both a specific patient and a group of patients identified by a common feature can be obtained.

As the object of the study, we used the data of 1997 electrocardiograms from 128 patients of the endocrinological hospital of the State Budgetary Healthcare Institution "City Clinical Hospital No. 52" DZM and the State Budgetary Healthcare Institution "Moscow Clinical Scientific Center named after A.S. Loginov »DZM with a stay of more than 5 days, as well as 741 electrocardiograms from 59 patients from different age groups without carbohydrate metabolism disorders (CFR), which was accepted as a control group.

We studied deviations in the forms of ECG electrocardiograms against the background of therapy during the period of inpatient follow-up in patients with carbohydrate metabolism disorders of the ICR using the CardioQVARK system (cardiomonitor and software), as well as differences in the forms of ECG in patients without ECG (control group) from the ECG forms of patients with NUO.

Was obtained the Fourier spectrum of the electrocardiogram (ECG) of a healthy patient (Fig. 1) and the Fourier spectrum of the electrocardiogram (ECG) with a sign of carbohydrate metabolism disorder (HCM) (Fig. 2).

The Fourier spectrum of an ECG with a sign of CHD is characterized by a pronounced expression of frequency harmonics, and the Fourier spectrum of an ECG of a healthy patient is characterized by the absence of a pronounced expression of frequency harmonics.

A cardiogram is a repetitive signal that is characterized by harmonics in the Fourier spectrum. The transfer of energy between the low frequency (LF) and high frequency (HF) ranges is characterized by the amount of noise between them. If the level of this noise is large, it means that there is a process of energy flow, either there or back. If the violation of synchronization of the process of energy flow is reduced, the authors consider this a sign of a violation of carbohydrate metabolism.

Auto return is the flow of energy between high and low frequency spectra back and forth.

You don't need several ECGs to compare that on one the energy flowed there, and on the other - in a slightly different way. If a cardiogram was taken once, then by the noise level one can conclude whether energy flows or does not flow.

It is impossible to determine where and where the overflow is coming from. Only a static picture can be observed at all times, even if several ECGs are compared.

Consecutively, on several ECGs, you can see: the LF part in one ECG decreases, the HF part rises. But in reality, they rise in different directions, i.e. some of the LF harmonics can increase, and some decrease.

For each spectrum, you can calculate the amount of chaos by harmonics. If a person is healthy, he will have a convex curve along the harmonics (on the Y-axis the amount of chaos, on the X-axis - the frequency), and for a patient with diabetes, a dip in the curve (concavity) is obtained, if prediabetes, the dip is small.

Not all states are fixed, more often the process of changing this state (transition state) is recorded. To record many conditions, you need to take a lot of cardiograms.

One ECG may indicate that the patient is in this state and this state is prevalent. Another ECG can tell that the patient is moving from one state to another. High frequencies rise or fall, this is only a separate frame of the entire process developing in time.

The functional activity of the brain is closely related to the use of glucose; therefore, a drop in the concentration of glucose in the bloodstream of the cerebral vessels necessitates an increase in the amplitude of the impulse that triggers the heart from the brain to increase the intensity of blood flow. This is due to a significant increase in the amplitude of the brain-generated pulse entering the sinus node, which leads to the formation of a more rigid Fourier shape of the harmonics of the ECG spectrum with a narrow pedestal, which is observed in patients with impaired carbohydrate metabolism. This, in turn, leads to a decrease in the degree of spectrum chaos, which can serve as a diagnostic sign of a violation of carbohydrate metabolism. This mode of cardiac activity corresponds to a certain solution of the Ginzburg – Landau chain of the order – disorder type. It should be noted that the magnitude of the feedback lag with a decrease in the glucose concentration in the cerebral bloodstream can be comparable with the reaction time of the brain upon the onset of hypoglycemic coma.

In addition to the circadian rhythm of the functioning of the pancreas, it also has a rhythm of functioning that is synchronous with the main frequency of heart contractions, equal to 1–2 Hz. Violation of this rhythm of the pancreas, as shown by a computer study of the dynamics of the spectrum of auto-return of FPU, leads to a decrease in the energy of the lowest frequency harmonics of the spectrum, which is observed in the spectra of patients with impaired carbohydrate metabolism.

Description of preferred embodiments of the invention

According to the invention, there is proposed a computerized method for non-invasive detection of disorders of carbohydrate metabolism by an electrocardiogram (ECG) of the heart. As indicated above, the heart is considered as an open nonlinear self-oscillating system, characterized by the Fermi-Pasta-Ulam (FPU) auto-return, while the ECG is a representation of the FPU auto-return. This method contains the following steps.

At least one electrocardiogram of at least the first lead is removed. One cardiogram is sufficient to detect the presence of an ICR violation.

Fourier transform of the specified at least one ECG is performed to obtain an amplitude Fourier spectrum for the specified at least one ECG (Fig. 1 and 2).

The Fourier amplitude spectrum of the specified at least one ECG is discretized to obtain a discrete view of the original form of the Fourier amplitude spectrum (DECG) of the specified at least one ECG (solid thick line in Fig. 1 or 2). The resulting DECG is used for computer processing, in which the DECG is analyzed to detect the presence / absence of a cardiac sign of a carbohydrate metabolism disorder (CVD) by establishing similarities / differences with the DECG of patients with CMD from the annotated ECG sample (Fig.3c) and the DECG of healthy patients from the annotated ECG sample (fig.3a)

The presence of similarity between the analyzed DECG and the DECG of patients with CMD and the absence of similarity with the DECG of healthy patients is a sign of a violation of carbohydrate metabolism.

In the specified method, the sampling step is carried out, for example, by constructing an envelope for the Fourier transformed at least one ECG in natural / averaged form (solid bold line in Fig. 1 and Fig. 2).

The sampling step can be performed by constructing the harmonic peak envelope for the Fourier transformed at least one ECG (bold dots in FIG. 1 and FIG. 2).

According to the invention, there is also proposed a computerized method for screening the population for the detection of persons with signs of carbohydrate metabolism disorders by their electrocardiogram (ECG) of the heart, which is considered as an open nonlinear self-oscillating system characterized by Fermi-Pasta-Ulam (FPU) auto-return, while the ECG is a representation auto return FPU. The screening method is performed as follows.

Specify targets for the sensitivity and specificity of screening, which are necessary to assess the effectiveness of the method.

The sensitivity of the study is the ability to distinguish between the two states of the analyzed patient: healthy and with pathology. Quantitatively, the sensitivity reflects the ratio of correct positive conclusions of the method to the total number of final diagnoses in a group of patients.

Specificity is the ability to reject disease, i.e. to state his absence where he really is not. Quantitatively, specificity reflects the proportion of negative conclusions in a group of people without a disease.

To determine the sensitivity and specificity, a group of people is examined for which it is known for each patient whether he is healthy or sick.

Sensitivity, specificity and predictive value of a positive result (PPV) are calculated using the formulas:

where A is the number of correct positive conclusions, B is the number of false positive conclusions, C is the number of false negative conclusions, D is the number of correct negative conclusions.

To calculate the sensitivity and specificity for detecting electrocardiographic abnormalities in patients with impaired carbohydrate metabolism, two groups were formed:

patients - inpatients (City Clinical Hospital No. 52 and MKSC) with type 2 diabetes mellitus, who had at least 2 ECGs (127 people);

healthy - staff of City Clinical Hospital No. 52, MKSC without diabetes and healthy from diabetes, who took at least 2 ECGs (59 people).

The results of applying the method by one ECG are presented in Table 1.

Table 1

Study Sick Healthy Recognized as sick 89 8 Recognized as healthy 38 51

Sensitivity and specificity can be significantly increased by deciding on multiple patient ECGs instead of just one. According to the results of the study, a table is compiled in which the number of cardiograms, the threshold value of the minimum number of cardiograms containing the ICO sign, and the corresponding indicators of sensitivity and specificity are set (Table 2).

Table 2 - Sensitivity and specificity of the method for determining CHD based on an ECG series

ECG number Threshold Sensitivity Specificity 1 1 70.08% 86.44% 2 1 83.46% 84.75% 2 2 58.28% 89.83% 3 1 88.18% 74.57% 3 2 76.38% 86.44% 3 3 48.03% 91.52%

Then, preparation for screening is carried out, for which the minimum required number of electrocardiograms (ECG) of one subject and the threshold value of the number of ECGs with signs of SMO for each age group of observation are determined according to table 2, taking into account the set target sensitivity and specificity.

And they screen the population to identify individuals with signs of a violation of carbohydrate metabolism. To do this, perform the following steps:

- the minimum required number of electrocardiograms (ECG) of at least the first derivation for one subject is removed, determined according to the table;

- carry out the Fourier transform of each captured ECG to obtain the amplitude Fourier spectrum for each captured ECG,

- discretization of the Fourier amplitude spectrum of each captured ECG is performed to obtain a discrete view of the initial form of the Fourier amplitude spectrum (DECG) of each captured ECG,

- use DECG for computer processing, in which DECG is analyzed to detect the presence / absence of a cardiac sign of a carbohydrate metabolism disorder (MCI) by establishing the similarity / difference with DECG of patients with MCI from an annotated sample of ECG and DECG of healthy patients from an annotated sample of ECG, while the presence of similarity in the analyzed DECG with DECG in patients with LUO and the lack of similarity with DECG in healthy patients is a sign of impaired carbohydrate metabolism.

And then a statistical analysis of the screening results is carried out, for which, from the minimum required number of ECGs of one subject, the number of ECGs with the OCD sign is determined and compared with the threshold value set during preparation for screening (according to Table 2), by which the presence of a carbohydrate metabolism disorder is judged the subject.

In this method, it is possible to additionally determine the severity of carbohydrate metabolism disorders by how much the number of ECGs with an ICG sign of one subject exceeds the threshold value, which results in a preliminary diagnosis of diabetes mellitus.

According to the invention, there is also proposed a computerized method for monitoring the level of carbohydrate metabolism disorders in a patient according to an electrocardiogram (ECG) of the heart, which is considered as an open nonlinear self-oscillating system characterized by Fermi-Pasta-Ulam (FPU) auto-return, wherein the ECG is a representation of the FPU auto-return. The monitoring method is performed as follows.

For each patient from the observed plurality of patients, the sequence of the captured plurality of his electrocardiograms (ECG) of at least the first lead marked with the patient identifiers is accumulated in the database. The specified sequence is obtained by adding each new recorded ECG of the patient, marked with the indicated identifiers, to his previously accumulated sequence of multiple ECGs throughout the observation period.

Fourier transform is performed for each ECG of the accumulated set of ECGs of the patient to obtain the amplitude Fourier spectrum for each ECG of the accumulated set of ECGs.

The Fourier amplitude spectrum of each accumulated ECG is discretized to obtain a discrete form of the Fourier amplitude spectrum (DECG) of each ECG of the accumulated set of ECGs (bold line in Fig. 1 and Fig. 2).

The resulting DECG is used for computer processing, in which the DECG is analyzed to detect the presence / absence of a cardiac sign of a carbohydrate metabolism disorder (CVD) by establishing the similarity / difference with the DECG of patients with CMD from an annotated sample of ECG and DECG of healthy patients from an annotated ECG sample, while the presence of similarity analyzed DECG with DECG of patients with MCI and the lack of similarity with DECG of healthy patients is a sign of impaired carbohydrate metabolism.

Further, the monitoring of the patient's condition over time can be carried out, according to one embodiment, by the change in time of the integral indicator of the carbohydrate metabolism disorder ICR, or, according to another embodiment, and by the change in the frequency of occurrence of signs of ICR in the patient's cardiograms.

When monitoring the patient's condition, the degree of difference between the analyzed DECG and the DECG of healthy people is calculated based on the change in time of the integral indicator of the DECG and is used as an integral indicator of disorders of carbohydrate metabolism.

The time change in the level of carbohydrate metabolism disturbance in the observed patient is estimated by the time change in the integral indicator of the carbohydrate metabolism disturbance.

And the prediction of the level of carbohydrate metabolism disturbance of the observed patient is carried out on the basis of the "observation history", which is the registered changes in time of the previous values of the integral indicator of the carbohydrate metabolism disturbance.

When monitoring for a change in the frequency of occurrence of signs of ICR in the patient's cardiograms, all ECG series are subdivided into two classes: containing the sign of CSD and those not containing the sign of CSD (Figs. 4a, 5a, 6a, 7a).

Summarize the statistical results of the analysis of the presence of a sign of CVD in a series of N ECGs of the patient. Depending on the frequency of manifestation of the ICD sign in the series, a conclusion is made about the severity of the already existing disorders of carbohydrate metabolism in the patient, up to a preliminary diagnosis of diabetes mellitus.

The dynamics of the increase - decrease in the frequency of occurrence of an ECG with an ICR sign is monitored on the basis of observations of the entire array of previously collected ECGs of the patient by comparing the frequency of occurrence of an ECG with an ICG sign for the previous and subsequent observation periods: days, weeks, months, years.

In diabetes, coronary heart disease progresses rapidly and is often the cause of a heart attack, therefore, non-invasive detection of carbohydrate metabolism disorders by patients on their own using an electrocardiogram in various conditions is important.

ECG diagnostics helps in an emergency to find out the cause of a heart attack in a matter of minutes and provide targeted assistance in a timely manner.

Below are examples of monitoring the level of impaired carbohydrate metabolism in patients according to the electrocardiogram (ECG) of the heart, which is considered as an open nonlinear self-oscillating system, characterized by Fermi-Pasta-Ulam (FPU) auto-return, while the ECG is a representation of the FPU auto-return.

In particular, we considered a patient with an established diagnosis of type 2 diabetes mellitus and patients without an established diagnosis of diabetes mellitus, ie. healthy patients.

EXAMPLE 1

Patient A, age 73. Diagnosis: type 2 diabetes mellitus for 25 years. There are no associated chronic diseases. The general condition according to the results of the examination is satisfactory.

The electrocardiogram (ECG) of the heart was monitored. The ECG was taken from 5 to 20 times a week from September 2017 to August 2019 (Fig.4a).

The accumulation in the database of the sequence of the captured plurality of his electrocardiograms (ECG) of at least the first derivation, marked with patient identifiers, was carried out. This sequence was obtained by adding each new recorded ECG of the patient to his previously accumulated sequence of multiple ECGs throughout the observation period.

Fourier transform was performed for each of the captured multiple ECGs to obtain the amplitude Fourier spectrum for each of the captured multiple ECGs.

Produced discretization of the amplitude Fourier spectrum of each of the captured set of ECGs to obtain a discrete view of the original form of the amplitude Fourier spectrum (DECG) of each of the captured set of ECGs (examples of figure 1 and figure 2).

As a result, all ECG series are subdivided into two classes: those containing the ICD sign and those not containing the ICG sign (Fig. 4a).

Computer processing was carried out, in which DECG was analyzed to detect the presence / absence of cardiac sign of carbohydrate metabolism disorder (CVD) by establishing the similarity / difference with DECG of patients with CMD from an annotated sample of ECG and DECG of healthy patients from an annotated sample of ECG.

The degree of difference between the analyzed DECG and the DECG of healthy people was calculated and used as an integral indicator of impaired carbohydrate metabolism.

The change over time in the level of impaired carbohydrate metabolism of the patient was assessed by the change in time of the integral indicator of impaired carbohydrate metabolism.

The prediction of the level of carbohydrate metabolism disorder of patient A was carried out on the basis of the "observation history", which is the registered changes in time of the previous values of the integral indicator of the carbohydrate metabolism disorder.

The total share of ECGs with cardiac signs of ICR is 99% and without ICG - 1% for the entire observation period (Fig. 4c).

The dynamics of the increase - decrease in the frequency of occurrence of an ECG with a symptom of ICR was monitored on the basis of observations of the entire array of previously collected ECGs of the patient by comparing the frequency of occurrence of an ECG with a sign of ICG for the previous and subsequent periods of observation: days, weeks, months, years.

There is a trend (dashed line) that for patient A with established type 2 diabetes mellitus, there is no dynamics of the frequency of manifestation of the ICD sign (Fig. 4a), which may indicate a positive trend in relation to the normalization of carbohydrate metabolism.

It follows from the above that for a patient with established type 2 diabetes mellitus, there is no dynamics in the frequency of manifestation of the symptom of SMD (Fig. 4a).

EXAMPLE 2.

Patient B, age 73. Diagnosis: diabetes is not diagnosed. There are no associated chronic diseases. The general condition according to the results of the examination is satisfactory.

The electrocardiogram (ECG) of the heart was monitored. The ECG was taken from 5 to 20 times a week from February 2016 to August 2019 (Fig.5a).

The accumulation in the database of the sequence of the captured plurality of his electrocardiograms (ECG) of at least the first derivation, marked with patient identifiers, was carried out. This sequence was obtained by adding each new recorded ECG of the patient to his previously accumulated sequence of multiple ECGs throughout the observation period.

Fourier transform was performed on each of the captured multiple ECGs to obtain the amplitude Fourier spectrum for each of the captured multiple ECGs.

Produced discretization of the amplitude Fourier spectrum of each of the captured set of ECGs to obtain a discrete view of the original form of the amplitude Fourier spectrum (DECG) of each of the captured set of ECGs (examples of figure 1 and figure 2).

As a result, all ECG series are subdivided into two classes: those containing the ICD sign and those not containing the ICG sign (Fig. 5a).

Computer processing was carried out, in which DECG was analyzed to detect the presence / absence of cardiac sign of carbohydrate metabolism disorder (CVD) by establishing the similarity / difference with DECG of patients with CMD from an annotated sample of ECG and DECG of healthy patients from an annotated sample of ECG.

The degree of difference between the analyzed DECG and the DECG of healthy people was calculated and used as an integral indicator of impaired carbohydrate metabolism.

The change over time in the level of impaired carbohydrate metabolism of the patient was assessed by the change in time of the integral indicator of impaired carbohydrate metabolism.

The prediction of the level of carbohydrate metabolism disorder in patient B was carried out on the basis of the "observation history", which is the recorded changes in time of the previous values of the integral indicator of the carbohydrate metabolism disorder.

The total share of ECGs with cardiac signs of ICR is 47% and without ICG - 53% for the entire observation period (Fig. 5c).

The dynamics of the increase - decrease in the frequency of occurrence of an ECG with a symptom of ICR was monitored on the basis of observations of the entire array of previously collected ECGs of the patient by comparing the frequency of occurrence of an ECG with a sign of ICG for the previous and subsequent periods of observation: days, weeks, months, years.

For patient B, in whom diabetes mellitus is not diagnosed, there is a trend (dashed line) for a decrease in the frequency of manifestation of the symptom of ICD, which may indicate a positive trend in relation to the normalization of carbohydrate metabolism.

Thus, this is an example of observation for a patient with an increase in the incidence of signs of CMD.

EXAMPLE 3.

Patient B, 44 years old. Diagnosis: diabetes is not diagnosed. There are no associated chronic diseases. The general condition according to the results of the examination is satisfactory.

The electrocardiogram (ECG) of the heart was monitored. The ECG was taken from 5 to 20 times a week from April 2017 to August 2019 (Fig.6a).

The accumulation in the database of the sequence of the captured plurality of his electrocardiograms (ECG) of at least the first derivation, marked with patient identifiers, was carried out. The specified sequence was obtained by adding each new recorded ECG of the patient to his previously accumulated sequence of multiple ECGs throughout the observation period.

Fourier transform was performed for each of the captured multiple ECGs to obtain the amplitude Fourier spectrum for each of the captured multiple ECGs.

Produced discretization of the amplitude Fourier spectrum of each of the captured set of ECGs to obtain a discrete view of the original form of the amplitude Fourier spectrum (DECG) of each of the captured set of ECGs (examples of figure 1 and figure 2).

As a result, all ECG series are subdivided into two classes: containing the ICD sign and those not containing the ICG sign (Fig. 6a).

Computer processing was carried out, in which DECG was analyzed to detect the presence / absence of cardiac sign of carbohydrate metabolism disorder (CVD) by establishing the similarity / difference with DECG of patients with CMD from an annotated sample of ECG and DECG of healthy patients from an annotated sample of ECG.

The degree of difference between the analyzed DECG and the DECG of healthy people was calculated and used as an integral indicator of impaired carbohydrate metabolism.

The change over time in the level of impaired carbohydrate metabolism of the patient was assessed by the change in time of the integral indicator of impaired carbohydrate metabolism.

The prediction of the level of carbohydrate metabolism disorder in patient B was carried out on the basis of the "observation history", which is the recorded changes in time of the previous values of the integral indicator of the carbohydrate metabolism disorder.

The total share of ECGs with cardiac signs of ICR is 29% and without ICG - 71% for the entire observation period (Fig. 6c).

The dynamics of the increase - decrease in the frequency of occurrence of an ECG with a symptom of ICR was monitored on the basis of observations of the entire array of previously collected ECGs of the patient by comparing the frequency of occurrence of an ECG with a sign of ICG for the previous and subsequent periods of observation: days, weeks, months, years.

For patient B, in whom diabetes mellitus is not diagnosed, there is a trend (dashed line) for a decrease in the frequency of manifestation of the ICD sign, which may indicate a positive trend in relation to the normalization of carbohydrate metabolism.

Thus, this is an example of observation for a patient with a decrease in the incidence of symptoms of CMD.

EXAMPLE 4

Patient D, age 34. Diagnosis: diabetes is not diagnosed. There are no associated chronic diseases. The general condition according to the examination results is satisfactory.

The electrocardiogram (ECG) of the heart was monitored. The ECG was taken from 5 to 20 times a week from April 2017 to August 2019 (Fig.7a).

The accumulation in the database of the sequence of the captured plurality of his electrocardiograms (ECG) of at least the first derivation, marked with patient identifiers, was carried out. This sequence was obtained by adding each new recorded ECG of the patient to his previously accumulated sequence of multiple ECGs throughout the observation period.

Fourier transform was performed on each of the captured multiple ECGs to obtain the amplitude Fourier spectrum for each of the captured multiple ECGs.

Produced discretization of the amplitude Fourier spectrum of each of the captured set of ECGs to obtain a discrete view of the original form of the amplitude Fourier spectrum (DECG) of each of the captured set of ECGs (examples of figure 1 and figure 2).

As a result, all ECG series are subdivided into two classes: those containing the ICD sign and those not containing the ICG sign (Fig. 7a).

Computer processing was carried out, in which DECG was analyzed to detect the presence / absence of cardiac sign of carbohydrate metabolism disorder (CVD) by establishing the similarity / difference with DECG of patients with CMD from an annotated sample of ECG and DECG of healthy patients from an annotated sample of ECG.

The degree of difference between the analyzed DECG and the DECG of healthy people was calculated and used as an integral indicator of impaired carbohydrate metabolism.

The change over time in the level of impaired carbohydrate metabolism of the patient was assessed by the change in time of the integral indicator of impaired carbohydrate metabolism.

The prediction of the level of carbohydrate metabolism disorder in patient B was carried out on the basis of the "observation history", which is the recorded changes in time of the previous values of the integral indicator of the carbohydrate metabolism disorder.

The total proportion of ECGs with cardiac signs of ICR is 9% and without ICG - 91% for the entire observation period (Fig. 7c).

The dynamics of the increase - decrease in the frequency of occurrence of an ECG with a symptom of ICR was monitored on the basis of observations of the entire array of previously collected ECGs of the patient by comparing the frequency of occurrence of an ECG with a sign of ICG for the previous and subsequent periods of observation: days, weeks, months, years.

For patient D, in whom diabetes mellitus is not diagnosed, there is a trend (dashed line) for a decrease in the frequency of manifestation of a symptom of ICD, which may indicate a positive trend in relation to the normalization of carbohydrate metabolism.

Thus, this is an example of observation for a patient with a decrease in the incidence of symptoms of CMD.

Claims (37)

1. A computerized method for non-invasive detection of carbohydrate metabolism disorders by the electrocardiogram (ECG) of the heart, which is considered as an open nonlinear self-oscillating system characterized by Fermi-Pasta-Ulam (FPU) auto-return, while the ECG is a representation of the auto-return of the FPU, containing the stages at which: - carry out the removal of at least one electrocardiogram of at least the first lead; - carry out the Fourier transform of the specified at least one ECG to obtain the amplitude Fourier spectrum for the specified at least one ECG; - produce discretization of the Fourier amplitude spectrum of the specified at least one ECG to obtain a discrete form of the original form of the Fourier amplitude spectrum (DECG) of the specified at least one ECG; - use the obtained DECG for computer processing, in which - analyze DECG to detect the presence / absence of a cardiac sign of carbohydrate metabolism disorder (MCI) by establishing similarity / difference with DECG of patients with MCI from an annotated sample of ECG and DECG of healthy patients from an annotated sample of ECG, while the presence of similarity of the analyzed DECG with DECG of patients with MCI and the lack of similarity with DECG in healthy patients is a sign of impaired carbohydrate metabolism. 2. The computerized method of claim 1, wherein the sampling step is performed by constructing a Fourier envelope of the transformed at least one ECG in natural / averaged form. 3. The computerized method of claim 1, wherein the sampling step is performed by constructing a harmonic peak envelope for the Fourier transformed at least one ECG. 4. A computerized method of screening the population to identify persons with signs of carbohydrate metabolism disorders by their electrocardiogram (ECG) of the heart, which is considered as an open nonlinear self-oscillating system, characterized by Fermi-Pasta-Ulam (FPU) auto-return, while the ECG is a representation of the auto-return of the FPU, containing the stages at which: - set targets for the sensitivity and specificity of screening; - carry out preparation for screening, for which it is determined according to the table, taking into account the set target indicators of sensitivity and specificity, the minimum required number of electrocardiograms (ECG) of one subject and the threshold value of the number of ECGs with signs of SME for each age group of observation, upon exceeding which the presence of ICR in patients of the group, and - conduct screening of the population to identify persons with signs of carbohydrate metabolism disorders, for which - the minimum required number of electrocardiograms (ECG) of at least the first derivation for one subject is removed, determined according to the table; - carry out the Fourier transform of each captured ECG from a certain number of ECGs to obtain an amplitude Fourier spectrum for each captured ECG; - produce discretization of the Fourier amplitude spectrum of the specified each captured ECG to obtain a discrete form of the original form of the Fourier amplitude spectrum (DECG) of the specified each captured ECG; - use DECG for computer processing, in which - analyze DECG to detect the presence / absence of a cardiac sign of carbohydrate metabolism disorder (MCI) by establishing similarity / difference with DECG of patients with MCD from an annotated sample of ECG and DECG of healthy patients from an annotated sample of ECG, while the presence of similarity of the analyzed DECG with DECG of patients with MCD and lack of similarity with DECG of healthy patients is a sign of a violation of carbohydrate metabolism; - a statistical analysis of the screening results is carried out, for which, from the minimum required number of ECGs of one subject, the number of ECGs with an ICG sign is determined and compared with a predetermined threshold value, by which the presence of a carbohydrate metabolism disorder in the subject is judged. 5. The computerized method according to claim 4, wherein additionally - the severity of carbohydrate metabolism disorders is judged by how much the number of ECGs with a sign of CSD in one subject exceeds the threshold value, which results in a preliminary diagnosis of diabetes mellitus. 6. A computerized method for monitoring the level of impaired carbohydrate metabolism of a patient according to the electrocardiogram (ECG) of the heart, which is considered as an open nonlinear self-oscillating system characterized by Fermi-Pasta-Ulam (FPU) auto-return, while the ECG is a representation of the FPU auto-return containing : - for each patient of the observed patients, a sequence of electrocardiograms (ECG) of at least the first lead marked with patient identifiers is accumulated in the database, the specified sequence is obtained by adding each new recorded ECG of the patient, marked with the indicated identifiers, to his previously accumulated ECG sequence on throughout the observation period; - carry out the Fourier transform of each of the captured ECGs to obtain the amplitude Fourier spectrum for each of the captured ECGs; - discretization of the Fourier amplitude spectrum of each of the captured ECGs is performed to obtain a discrete view of the original form of the Fourier amplitude spectrum (DECG) of each of the captured ECGs; - use the obtained DECG for computer processing, in which - analyze DECG to detect the presence / absence of a cardiac sign of carbohydrate metabolism disorder (MCI) by establishing similarity / difference with DECG of patients with MCD from an annotated sample of ECG and DECG of healthy patients from an annotated sample of ECG, while the presence of similarity of the analyzed DECG with DECG of patients with MCD and lack of similarity with DECG of healthy patients is a sign of a violation of carbohydrate metabolism; - the degree of difference between the analyzed DECG and the DECG of healthy people is calculated and used as an integral indicator of the violation of carbohydrate metabolism; - the change in time of the level of carbohydrate metabolism disturbance in the observed patient is assessed by the change in time of the integral indicator of the carbohydrate metabolism disturbance; - prediction of the level of carbohydrate metabolism disorder of the observed patient is carried out on the basis of the "observation history", which is the registered changes in time of the previous values of the integral indicator of the carbohydrate metabolism disorder. 7. A computerized method for monitoring the level of impaired carbohydrate metabolism of a patient according to the electrocardiogram (ECG) of the heart, which is considered as an open nonlinear self-oscillating system characterized by Fermi-Pasta-Ulam (FPU) auto-return; : - for each patient of the observed patients, a sequence of electrocardiograms (ECG) of at least the first lead marked with patient identifiers is accumulated in the database, the specified sequence is obtained by adding each new recorded ECG of the patient, marked with the indicated identifiers, to his previously accumulated ECG sequence on throughout the observation period; - carry out the Fourier transform of each of the captured ECGs to obtain the amplitude Fourier spectrum for each of the captured ECGs; - discretization of the Fourier amplitude spectrum of each of the captured ECGs is performed to obtain a discrete view of the original form of the Fourier amplitude spectrum (DECG) of each of the captured ECGs; - use the obtained DECG for computer processing, in which - analyze DECG to detect the presence / absence of a cardiac sign of carbohydrate metabolism disorder (MCI) by establishing similarity / difference with DECG of patients with MCD from an annotated sample of ECG and DECG of healthy patients from an annotated sample of ECG, while the presence of similarity of the analyzed DECG with DECG of patients with MCD and lack of similarity with DECG of healthy patients is a sign of a violation of carbohydrate metabolism; - the time change in the level of carbohydrate metabolism disturbance in the observed patient is assessed by the frequency of occurrence of cardiograms with a sign of CHD - prediction of the level of carbohydrate metabolism disturbance of the observed patient is carried out on the basis of the "observation history", which is the recorded changes in time of the previous values of the frequency of occurrence of cardiograms with a sign of ICR.

Images