Cardiac Monitoring Method

Description

[0001] The invention relates to a cardiac monitoring method according to the preamble of claim 1.

[0002] Coronary heart disease, or CHD for short, is one of the most common illnesses among patients. The current prevalence in Germany (approximately 30% in men, approximately 15% in women) demonstrates a great need for devices that can quickly detect such problems and provide people or patients with early treatment or advice. Another major area of ​​cardiac disturbances is arrhythmia, or cardiac arrhythmia. The human heart has a complex excitation and conduction system (excitation system), which drives cardiac movements. Due to the division of the heart into ventricles and atria, it is necessary to regulate blood flow through muscle contractions. In connection with the rhythmic contraction of the heart muscle, the excitation system consists of special muscle cells that are capable of spontaneously depolarizing and repolarizing themselves.This allows a rhythm primarily generated by the sinus node to be maintained. A disruption of this excitation system impairs the heart's pumping capacity, resulting in supply bottlenecks. Not all symptoms are life-threatening, but if they occur frequently, they can be good indicators of impending life-threatening conditions.

[0003] The reference standard for detecting coronary heart disease is coronary angiography, although this falls outside the scope of early detection of coronary heart disease. This involves inserting a catheter into the heart to perform, among other things, aortic pressure measurement. Contrast media is also used to visualize the heart's coronary arteries. Accordingly, this is an invasive procedure with associated health risks. Alternatives include CT angiography and MRI. These alternatives also use contrast media, with associated health risks. These procedures are disadvantageously associated with high complexity and expense.

[0004] Compared to the gold standard of coronary diagnostics, coronary angiography, with a specificity and sensitivity of 100%, non-invasive procedures such as resting ECG and resting echocardiography demonstrate significantly lower sensitivity and specificity (less than 30% and 70%, respectively). These only increase under stress, but both require the presence of a physician. Furthermore, the results of stress echocardiography are also examiner-dependent. In addition, both of these procedures require an additional time of at least 15–30 minutes. Other non-invasive procedures such as myocardial perfusion synthography (MPS), coronary computed tomography (CCT), or cardiac magnetic resonance imaging (MRI) require considerable equipment and personnel resources, the presence of a physician, and are also significantly expensive.

[0005] An ECG can be performed for a non-invasive examination of patients. Modern software systems offer physicians many tools in this field, at least to quickly perform patient measurements. However, the evaluation of an ECG depends significantly on the experience of the treating physician and is therefore uncertain. Furthermore, overburdening physicians can lead to misdiagnoses. To solve this problem, automated measurement systems have been established in the field of ECG analysis. However, these automated systems are limited to measuring ECGs and thus map experience or defined parameters (e.g., QT interval or ST elevation) into the software. Thus, only the uncertainty caused by the human factor has been removed.

[0006] ECG parameters, especially at rest, are generally insufficient for detecting coronary heart disease. For this reason, a stress ECG is performed on patients to enable early detection of coronary heart disease. This requires that the patient is subjected to physical exertion. This provokes symptoms that could lead to a heart attack. This always requires the presence of a doctor during such a measurement. The disadvantage of this procedure lies in the associated risks. Aside from the danger to the patient, not everyone can undergo such tests. Especially in high-risk groups (diabetes, obesity, old age), people are not always able to achieve the necessary stress due to a lack of sufficient fitness. As an alternative, stress can be achieved with medication and then examination using echocardiography (stress echo) using ultrasound.However, the echo is more suitable for the examination of movement disorders and also does not offer the reliability of an automatic evaluation.

[0007] With currently available ECG devices, it is not possible to detect early signs of coronary heart disease or to present them to the physician in a suitable form. All devices known to date have in common that they display multiple leads simultaneously, usually with 3 to 12 channels. The display software usually allows the view to be configured, which allows a physician to concentrate on individual leads and evaluate them according to known guidelines. Suspicion of an impending myocardial infarction (the result of untreated coronary heart disease) arises, among other things, when ST segment elevations >= 0.1 mV in at least two extremity leads and >= 0.2 mV in two adjacent chest wall leads. However, the absence of this elevation does not mean that a myocardial infarction can be ruled out, because the majority of myocardial infarctions are non-elevation myocardial infarctions (NSTEMI).

[0008] In the article "R and T wave relationships in the lead I electrocardiogram" by Bloomfield DK et al. (JOURNAL OF ELECTROCARDIOLOGY, ELSEVIER SCIENCE, XX, Vol. 2, No. 2, April 1, 1969 (1969-04-01), pages 167-170, XP026701536, ISSN: 0022-0736, DOI: 10.1016 / S0022-0736(69)81021-6), the relationships between R and T waves in the lead I electrocardiogram are described. A significant correlation between R and T wave amplitudes was demonstrated, with the normal R / T ratio being determined to be 4.0 ± 1.6. When obvious diagnoses such as myocardial infarction, right bundle branch block, and pericarditis are added to the abnormal electrocardiograms detected by the R / T ratio, the detection rate of lead I screening improves to 84% of abnormalities. This suggests that lead I electrocardiograms play a significant role in mass screening.

[0009] EP 86 429 B1 discloses a method for cardiogoniometry that belongs to the class of vector cardiometric methods. This method allows for very precise determination of the sum vector representing the heart's electrical field and represents it in a Cartesian coordinate system aligned with the heart's longitudinal axis (rather than the body axis). For the diagnosis of heart disease, the potentials reached at their maximum during depolarization and repolarization (R and T loops) of the heart muscle are used, as are the directions of the corresponding two potential-specific vectors and the solid angle between the two specific vectors. This means that for diagnostic purposes, vectors with their magnitude and, separately, their directions and individual components are parameterized in Cartesian coordinates, spherical coordinates, and cylindrical coordinates.

[0010] The derivation points of cardiogoniometry according to EP 86 429 B1, between which potential differences are measured, are: E1 corresponding to the point for tension V4 (according to Wilson), E2 sagittal to E1 and corresponding to the point for tension V8 (according to Wilson), E3 perpendicular to the line E1-E2, above E1 at a distance of 0.7 times the distance between E1 and E2, E4 horizontal from E3 to the right side of the patient at a distance of 0.7 times the distance between E1 and E2.

[0011] A vector addition of the measured potential differences results in a space vector V or projections of the space vector V onto an x, y and z-axis, where the x-axis is parallel to the line segment E1-E2, and the y-axis is parallel to the line segment E1-E4 and where the z-axis is perpendicular to the plane spanned by E1, E2 and E4 (oblique sagittal plane).

[0012] The electrical action of the heart generates an electric field that changes over time. The spatial vector V determined according to EP 86 429 B1 is an approximation of the actual spatial sum vector of this field. The direction of the spatial vector corresponds to the field direction, and the length (magnitude or potential) of the spatial vector corresponds to the field strength. The three voltages X, Y, and Z [mV] represent the partial vectors or components of the spatial vector V in a three-dimensional Cartesian coordinate system, preferably the spatial vector determined using the cardiogoniometric method in the aforementioned coordinate system. The potential is calculated as the square root of (X 2< +Y 2< +Z 2< ).

[0013] From WO 99 / 36860, it is known that the temporal progression of the magnitude of the spatial vector V is divided into the ranges R+, R-, ST, T+, and T- by determining maxima and minima (zeros in the progression of the differential quotient). In these ranges, the periodically recorded vector magnitudes are summed, and the integral values ​​thus determined are further used for diagnostic purposes. Thus, loop-wise sums of vector magnitudes are parameterized.

[0014] Furthermore, WO 03 / 057031 A1 discloses a method for generating cardiometric parameters, which can be used in particular for diagnostic purposes in the examination of the human heart. Such a method is referred to in WO 03 / 057031 A1 as "vector cardiometry" and serves to generate cardiometric parameters from a vector that represents the electrical field of the heart and is determined, for example, using known vector cardiometric methods. The spatial vector is represented starting from a coordinate origin, with a spatial trajectory described by the vector tip and / or a spatial velocity of the vector tip moving along the trajectory being parameterized in virtual space around the coordinate origin. The parameters generated in this way, or the aforementioned deviations, are particularly suitable for diagnosing heart diseases or disorders.

[0015] The above-described methods for determining and displaying a spatial vector related to the heart using cardiogoniometry, particularly according to EP 86 429 B1, suffer from the disadvantage that electrodes or sensors must be placed on the human body in exact alignment with the four lead points E1-E4, where potential differences are measured. An incorrect position of an electrode, particularly at lead point E1, leads to a falsification of the measured values, rendering the resulting analysis unusable and thus preventing a reliable diagnosis for the heart being examined.

[0016] A further disadvantage of the technologies of the above-mentioned prior art documents is that they do not provide for automatic analysis of the three-dimensional signals obtained from a patient.

[0017] Accordingly, the object of the invention is to create a technology in the field of non-invasive cardiac monitoring which enables improved examination of patients in a shorter time and, in particular, also enables automatic analyses.

[0018] The above object is achieved by a method having the features specified in claim 1. Advantageous developments of the invention are defined in the dependent claims.

[0019] For monitoring the heart in a human, the present invention provides a method in which, based on ECG signals recorded from a patient's heart, a spatial vector is determined using vectorcardiography, wherein this spatial vector depicts the temporal progression of the sum vector of the heart's electrical field and has a direction corresponding to the field direction and a length corresponding to the potential. A quotient is formed from the areas swept over by a length of the spatial vector (= radius vector) as a function of time during the R wave and during the T wave, and this dimensionless and scalar quotient is subsequently subjected to further evaluation. During this further evaluation, the quotient formed from the area (R wave) / area (T wave) is in each case related to predetermined limit values ​​that are relevant for the clinical picture under investigation.This means that the said quotient, formed from the area (R-wave) / area (T-wave), is correlated or compared with at least one predetermined limit value, on the basis of which a diagnosis is then made or possible for the patient being examined with regard to a specific clinical picture (e.g. CHD and / or HRS and / or heart valve insufficiency of a patient). With this method, coronary heart disease (CHD) is detected or diagnosed for the heart being examined if the quotient formed from the areas swept over by the spatial vector during the R-wave and during the T-wave. Fläche R − Welle Fläche T − Welle = a is not in the interval [a 0,KHK ,a 1,KHK ] or outside this interval. This interval is defined by a lower limit a 0,KHK and an upper limit a 1,KHK. These limits a 0,KHK and a 1,KHK are determined depending on a training set. The lower and upper limits a 0 and a 1 are each determined using a training set of subject data, with the following steps: (i) Calculating the quotient area (R-wave) / area (T-wave) for all subjects to obtain ratio time series, (ii) Calculating a mean (µ) and a corresponding standard deviation (σ) from the ratio time series from step (i), (iii) Determining the lower limit a 0 taking into account the mean (µ) and the standard deviation (σ) from step (ii), using the relationship: a 0 = mean (µ) - standard deviation (σ), and / or Determining the upper limit a 1 taking into account the mean (µ) and the standard deviation (σ) from step (ii), using the relationship: a 1 = mean (µ) + standard deviation (σ). Wherein in step (ii) the mean and the corresponding standard deviation are calculated only for ratio time series from healthy patients ("- -").

[0020] Furthermore, it is pointed out that the indexing "CHD" of the lower and upper limit values ​​a 0,CHD and a 1,CHD for the present invention is to be understood in such a way that with the help of these limit values ​​the clinical picture of coronary heart disease ("CHD"), which is also known under the term ischemic heart disease ("IHD"), can be detected.

[0021] In the method according to the present invention just discussed, a cardiac arrhythmia (HRS) can also be detected for the examined heart if the quotient formed from the areas swept over by the space vector (10) during the R-wave and during the T-wave satisfies the condition Fläche R − Welle Fläche T − Welle < a 0 , HRS or Fläche R − Welle Fläche T − Welle > a 1 , HRS fulfilled.

[0022] Accordingly, according to the last-discussed method of the present invention, it is possible to make an accurate initial diagnosis for the examined heart of a patient with regard to a possible presence of HRS by comparing the quotient formed from the areas swept by a length of the space vector (= radius vector) as a function of time during the R-wave and during the T-wave, respectively, with predetermined values ​​for the parameter "a".

[0023] Depending on whether the specific diagnosis of HRS or CHD is to be made, it is understood that the threshold values ​​a 0 and a 1 are suitable for detecting these clinical pictures and thus for comparison or correlation with the quotient formed from the area (R-wave) / area (T-wave), which is based on corresponding data. In the context of HRS, this can be expressed by the index "HRS" (i.e., a 0,HRS , a 1,HRS ); in the context of CHD, this can be expressed by the index "CHD" (i.e., a 0,CHD , a 1,CHD ).

[0024] If reliable data for the lower and upper limit values ​​a 0 , a 1 are available with regard to a clinical picture to be examined (e.g., HRS and / or CHD), these limit values ​​can be set to these predetermined values ​​before a patient is examined, depending on whether the presence of HRS or CHD is to be investigated. Alternatively, it is possible to use these limit values ​​with the help of a training set of data records from a large number of test subjects, as explained in detail below. With regard to the above-mentioned steps (i) to (iii), it is understood that these are to be assigned to a training session with which the lower and upper limit values ​​a 0 and a 1 can be determined with greater "robustness" in order to subsequently and based on this, make an actual diagnosis of a patient for the clinical picture under investigation with greater accuracy.

[0025] To clarify the above-mentioned calculation rule in step (iii), it should be noted again that the lower limit a 0 is calculated by subtracting the mean µ and the corresponding standard deviation σ. In contrast, the upper limit a 1 is calculated by adding the mean µ and the corresponding standard deviation σ.

[0026] With regard to the aforementioned determination of the lower and upper limit values ​​a 0 and a 1 , it is understood that the underlying data of the test subjects with which the ratio time series are formed according to step (i) each relate to the clinical picture to be investigated, HRS and / or CHD, or are relevant for this. For example, to investigate the presence of coronary heart disease ("CHD"), the limit values ​​a 0,CHD and a 1,CHD are determined only on the basis of data sets from test subjects for whom it can be reliably ruled out that they suffer from CHD or IHD. Mutatis mutandis, this also applies to the limit values ​​a 0,HRS and a 1,HRS when examining a patient for the presence of HRS.

[0027] By using only the data sets of healthy patients ("- -") to determine the lower and upper limit values, the advantage is achieved in comparison to the data sets of sick patients that the standard deviations of the mean values ​​of the healthy patients are generally smaller, which allows the detection of the clinical pictures or findings HRS and / or CHD with greater accuracy and reliability.

[0028] In an advantageous development of the invention, in particular with regard to all of the methods mentioned above, it can be provided that a T-shirt is used to record the ECG signals. The T-shirt has four sensors or electrodes that are assigned to a correct position of the four lead points on the patient's body. These electrodes can be suitably incorporated into the fabric of the T-shirt. In the same way as a T-shirt, a wearable or mobile sensor system can also be provided for the sensors or electrodes that are to be brought into contact with the patient's body, e.g. in the form of a chest strap or the like. It is important here that with such a wearable sensor system the individual sensors or electrodes are connected to one another to form a composite, thus achieving precise positioning at the respective lead points on the patient's body.Accordingly, there is no need to manually apply the individual sensors to the patient’s body.

[0029] The theoretical basis for the analysis principle according to the present invention is the so-called "vectorcardiography", which makes a significant contribution in the field of cardiac ischemia diagnostics as a non-invasive method.

[0030] In connection with the method according to the present invention, the representation of the electrophysiological properties of the heart is simplified compared to the complicated representation of the often used conventional ECG with its up to 12 leads. This is achieved by first converting the filtered ECG signals into orthogonalized measured values ​​based on vectorcardiography. In this process, the cardiac potentials are measured in an orthogonal system and summed vectorially. With such a three-dimensional measurement and localization of the cardiac potential, a multitude of new parameters can be created or taken into account, which allows, among other things, the detection of myocardial ischemia.

[0031] A key advantage of the invention is that it enables automated analysis using three-dimensional signals. This not only ensures a profound basis for analysis but also enables graphical processing of the heart's electrophysiological properties. Compared to previously used ECG measurements, which utilize up to 12 leads (= 12 dimensions), the otherwise complex representation is thus greatly simplified and visualized in an intuitive manner, thus reducing errors in analysis and evaluation with the present invention.

[0032] The measurement or non-invasive recording of ECG signals from the patient's heart is preferably performed in a resting state (i.e., without physical exertion) using four surface electrodes. This means that the ECG signals are recorded at a total of four lead points on the patient's body near the heart. The following display formats can be used for this measurement: Analog orthogonal projections X, Y, Z for checking the technical quality of the measurement; potential measured at different time points of the cardiac cycle; 2D and 3D loops for locating the potential in the heart, maximum vectors of atrial and ventricular depolarization and ventricular repolarization.

[0033] A central aspect of the present invention is that the orthogonalized measured variables are input into a system based on artificial intelligence (AI), preferably in the form of a neural network or a plurality of such neural networks. This AI system (or neural network) has been trained with previously known findings, and to the extent that these findings from comparison patients are stored in the AI ​​system. A diagnosis for the examined patient is created by functionally mapping the input orthogonalized measured variables via a neural network to values ​​between -1 and 1. This AI system has been trained with the aid of findings from comparison patients, which weights typical characteristics of the measured variables according to their relevance to the state of health.

[0034] With regard to the known diagnostic data of comparison patients on which the AI ​​system or neural network is based, it is specifically emphasized that there is clear knowledge of the health status ("healthy" or "ill") for these comparison patients. This information may, for example, have been obtained through a coronary angiography, which provides a clear diagnosis for the comparison patients ("healthy" or "ill"), or may be available from other sources. Based on this, the diagnostic data is classified and clustered (diagnosis as "healthy" or "ill").

[0035] In an advantageous development of the invention, the leads between the four respective lead points at which the ECG signals are recorded are viewed both in Cartesian coordinates and in spherical and cylindrical coordinates. Such a transformation into spherical coordinates enables the determination of the actual times of the physiological extreme points (i.e., the exact point of the R-peak or the T-wave) during a heartbeat.

[0036] The aforementioned method according to the present invention is a non-invasive, reproducible, rapid, and cost-effective diagnostic approach for detecting hemodynamically relevant stenoses of the coronary arteries at rest. The diagnosis is made via a computer-based, infinitesimal, three-dimensional calculation of the excitation processes of the mammalian heart based on a specific algorithm in conjunction with an AI system or a neural network, correlating the intrinsic blood supply and the specific spatial orientation of the myocardium in the dipole field as a function of time, starting from a defined point. With a method according to the invention, it is possible to achieve a sensitivity of 95 to over 99% and a specificity of 80 to over 90% for the examined patients. This approach is also referred to by the applicant as "cardisiography."

[0037] With regard to the present invention, it is specifically emphasized that its implementation is preferably carried out using a computer or comparable computing units. This applies to the method according to claim 1 with regard to the evaluation of the quotient formed from the area (R-wave) / area (T-wave) taking into account at least one predetermined limit value.

[0038] Various embodiments of the invention are described in detail below with reference to a schematically simplified drawing.

[0039] They show: Fig. 1 a simplified schematic diagram of a device for the early detection of the presence of CHD or HRS, Fig. 2 the view of a patient from the front ( Fig. 2a ) and from behind ( Fig. 2b ), Fig. 3 the view of a T-shirt from the front ( Fig. 3a ) and from behind ( Fig. 3b ), which is used in accordance with Fig. 1 can be used, Fig. 4, Fig. 7b each show the front view of a patient, to illustrate an orthogonal system (x, y, z) in which a space vector is mapped according to the present invention, Fig. 5 schematic simplified views of a first triangle ( Fig. 5a ) and a second triangle ( Fig. 5b ), on the basis of which, with respect to the orthogonal system according to Fig. 4 an angle correction is carried out, Fig. 6 a fundamentally simplified view of a system architecture within which the establishment of Fig. 1 is used, Fig. 7a a fundamentally simplified representation of a space vector which is formed during an activity of the human heart, Fig. 8 a flow chart with which a method which is not part of the invention is carried out, Fig. 9 a flow chart with which a method which is not part of the invention is carried out according to a further embodiment, Fig. 10 a fundamentally simplified representation of a 3D matrix ( Fig. 10a ) of measured values ​​or a standardized version thereof ( Fig. 10b ), with the statistics derived from the parameter time series or parameters, which in a procedure according to Fig. 8 or Fig. 9 is used, Fig. 11a-11dexample time series parameters that are used in a method according to Fig. 9 are applicable, Fig. 12exemplary statistical methods that can be used in a method according to Fig. 9 are applicable, Fig. 13 a flow chart with which a method which is not part of the invention is carried out, Fig. 14 a further flow chart with which a method which is not part of the invention is carried out according to a further embodiment, Fig. 15 the multiplication of a calculation matrix with a calculation vector according to a further method which is not part of the invention, Fig. 16 a basically simplified main axis transformation for the method of Fig. 15 , Fig. 17 a simplified illustration of the formation of a quotient from the areas swept by a length of the space vector (= radius vector) as a function of time during the R-wave and during the T-wave, and Fig. 18 a simplified illustration of a plurality of neural networks used in an application or implementation of the present invention.

[0040] The following are with reference to the Fig. 1-18 Preferred embodiments of cardiac monitoring methods and a device 10 used therefor are illustrated and explained in detail. Identical features in the drawings are provided with the same reference numerals. It is specifically noted here that representations in the drawings are merely simplified and, in particular, are shown without scale.

[0041] The facility 10, in Fig. 1 In principle, presented in a simplified manner, it is used in the examination of a patient 11 (cf. Fig. 2 ) for the early detection of coronary heart disease (CHD) or cardiac arrhythmia (HRS), and is not part of the invention. For this purpose, the device 10 comprises four sensors or electrodes S1, S2, S3, S4, a data filter 16, an evaluation device 18 and a system 20 based on artificial intelligence (AI). The evaluation device 18 is equipped with a memory element (not shown) so that measurement signals or values ​​can be stored therein, at least temporarily. The AI ​​system 20 can comprise at least one neural network 20N (cf. also Fig. 18 ) or have a plurality of such neural networks or be formed from such neural networks.

[0042] The sensors S1-S4 are connected to the device 10 via signaling (e.g., via a cable connection or a wireless radio link) in such a way that their measured values ​​first pass through the filter 16 and then reach the evaluation device 18. The evaluation device 18 is connected to the AI ​​system 20 via data technology in such a way that the measured variables, which are suitably processed by the evaluation device 18 or converted into orthogonalized measured variables based on vectorcardiography, as will be described separately below, can be input into the AI ​​system 20. This is done for the purpose of a diagnosis for the patient 11 being examined using the device 10.

[0043] Fig. 2 shows - for a better understanding of the invention - a patient 11, namely in a view from the front ( Fig. 2a ) and in a view from behind ( Fig. 2b ). A total of four lead points are provided on the body 14 of the patient 11, namely a first lead point E1, a second lead point E2, a third lead point E3 and a fourth lead point E4. The lead points E1, E3 and E4 are each located in the chest region of the patient 11, with the lead point E2 being located in the back region of the patient 11. With regard to these four lead points E1 to E4, it should be noted that the first sensor S1 is attached to the first lead point E1, the second sensor S2 to the second lead point E2, the third sensor S3 to the third lead point E3 and the fourth sensor S4 to the fourth lead point E4 on the body 14 of the patient 11. Potential differences are measured between each of these lead points, as explained separately below.For further details on the positions of these individual derivation points E1-E4 on the patient's body 14, reference is made to the disclosure according to EP 86 429 B1, the contents of which are hereby incorporated by reference in their entirety.

[0044] In the Fig. 3 A simplified T-shirt 22 is shown. The above-mentioned sensors S1-S4 of the device 10 can be integrated into the T-shirt 22, for example by being woven into its textile structure. Such a T-shirt 22 has the advantage that a patient 11 simply puts on or pulls on this T-shirt 22 in preparation for an examination, whereby the sensors S1-S4 integrated into the T-shirt 22 then automatically move into their intended position adjacent to the four lead points E1-E4. The use of such a T-shirt 22 eliminates the time-consuming and potentially error-prone manual application of the individual sensors S1-S4 to the body 14 of the patient 11. As an alternative to the T-shirt 22, a chest strap (not shown) to which the sensors or electrodes S1-S4 are attached can also be used.

[0045] To examine a patient 11 or to obtain a sample data set for training the AI ​​system 20, the four sensors S1-S4 are positioned at the assigned four lead points E1-E4 on the human body 14. Subsequently, while the patient 11 is at rest, ECG signals are recorded at the heart 12 of the patient 11 using the positioned sensors S1-S4. The ECG signals are then appropriately filtered by the filter 16 and subsequently transformed in the evaluation device 18 into orthogonalized measured variables (in the x, y, z axes) according to an orthogonal system.

[0046] In the Fig. 4 the axes x, y, z are illustrated according to an orthogonal system, with respect to the body 14 of a patient 11 and his heart 12. In this regard, reference is also made to the Fig. 5 referred to, whereby in Fig. 5a schematically illustrates a first triangle 31 formed between the derivation points E1, E3 and E4, and in Fig. 5b a second triangle 32 is illustrated, which is chosen between the derivation points E1, E2, and E4. The meaning of these two triangles 31, 32 will be explained separately below.

[0047] As already explained, potential differences are recorded between the individual lead points E1-E4. Specifically, these are an anterior lead A between the first lead point E1 and the fourth lead point E4, a dorsal lead D between the second lead point E2 and the fourth lead point E4, a horizontal lead H between the third lead point E3 and the fourth lead point E4, a vertical lead V between the first lead point E1 and the third lead point E3, and finally an inferior lead I between the first lead point E1 and the second lead point E2. From the first lead point E1, the third lead point E3 and the fourth lead point E4, as shown in Fig. 5a - a first triangle 31 is formed, wherein from the first derivative point E1, the second derivative point E2 and the fourth derivative point E4 - as shown in the illustration in Fig. 5b - a second triangle 32 is formed. The above-mentioned derivatives are also shown with their designations A, D, H, V and I in the representations of Fig. 5a und Fig. 5b shown. For further information regarding these derivations, reference may be made to the content of EP 86 429 B1.

[0048] The meaning of the first and second triangles 31, 32 is explained separately elsewhere below in connection with a so-called "correction method" according to the present invention.

[0049] When performing an ECG measurement, the electrical measured values ​​of the above-mentioned leads A, D, H, V and I are transferred to the device 10 and are further processed therein, as already explained above. With reference to the Fig. 6 Further details of the device 10 and its integration into an overall architecture for implementing the present invention are explained below. In detail: The architecture according to Fig. 6 provides the following components: sensor device 100, data recorder 102, Cardisio®< device 106, and server 113. The data recorder 102 comprises a signal receiver 103, a signal converter 104, and a signal memory 105. The Cardisio® device 106 comprises a signal reader 107, a vector data generator 108, a vector data memory 109, a vector data evaluator 110, a vector data synchronizer 111, and a vector data display 112. The server 113 comprises a vector data memory 114, a vector data analyzer 115, and a vector data evaluation memory 116.

[0050] The sensor device 100 is formed by the above-mentioned four sensors or electrodes S1-S4 of the device 10.

[0051] The sensor device 100 and the data recorder 102 form the components of the device 10 that serve to measure and record the ECG signals. These receive, process, and appropriately transform the analog ECG signals into digital signals. As already explained, the digital signals can be stored, at least temporarily, in the memory element of the evaluation device 18—here in the form of the signal memory 105.

[0052] The Cardisio device 106 reads the digital signals from the data recorder 102 for the purpose of determining vector data based on the digital signals using the vector data generator 108. The vector data generated in this way is then stored in the vector data memory 109. Based on this, the vector data evaluator 110 generates a representation of this vector data, e.g., in the form of a three-dimensional curve, which representation is then shown or visualized using the vector data display 112.

[0053] The architecture according to Fig. 6 further illustrates that within the Cardisio device 106, the vector data evaluator 110 is signal-connected to the vector data synchronizer 111, and the latter is connected to the server 113 via a signal path or connection. This allows the generated vector data to be initially read into the vector data memory 114 on the server 113 via the vector data synchronizer 111. Subsequently, the vector data analyzer 115 can be used to perform a targeted evaluation of larger data volumes or the generated vector data and to perform statistical analyses in this regard. Finally, the manual and / or automatic evaluations of the data sets are stored in the vector data evaluation memory 116, from which these evaluations can be uploaded back to the device 106, e.g., for display on or with the vector data display 112.

[0054] For those based on the Fig. 6 explained components and parts of the sensor device 100, the data recorder 102, the Cardisio device 106 and the server 113, it is understood that these are each parts of the device 10 (cf. Fig. 1 ). In particular, the vector data evaluation memory 116 is part of the AI ​​system 20 or a neural network 20 N , wherein already known findings data from comparison patients are stored therein, for whom a clear diagnosis ("healthy" or "ill") is known.

[0055] Based on the ECG measurement data recorded on the heart 12 of a patient 11 using the four sensors S1-S4, the evaluation device 18 can generate a spatial vector 24 that maps the electrical activity of the heart 12. Specifically, this spatial vector 24 represents the temporal profile of the sum sector of the electrical field of the heart 12 and has a direction corresponding to the field direction and a length corresponding to the potential. Such a spatial vector 24 is illustrated, for example, in the Fig. 7a which is preferably formed in an orthogonal system with the axes x, y, z. Fig. 7b shows again - in the same way as Fig. 4 - in a simplified manner the heart 12 of a patient 11, in connection with the axes x, y, z.

[0056] Below, a method which is not part of the invention is described with reference to the Fig. 8 which shows a flowchart of steps of such a method. In detail: At the beginning of the method, the sensors S1-S4 of the device 10 explained above (cf. Fig. 1 ) is attached to the upper body of a patient 11, namely in correspondence with the four derivation points E1, E2, E3 and E4 (cf. Fig. 2a, Fig. 2b ). For this purpose, the T-shirt 22 from Fig. 3 In this way, ECG signals are recorded non-invasively at the heart 12 of the patient 11, namely in the resting state of the patient 11. This corresponds to step (i) of the method according to Fig. 8 .

[0057] Subsequently, in a step (ii) of the procedure, Fig. 8 the recorded ECG signals are filtered, namely, as described with reference to Fig. 1 explained, by filter 16. Such filtering serves to eliminate high-frequency noise and low-frequency interference (e.g., caused by the patient's breathing). Examples of filter types are notch filters, high-pass filters, and Sarvitzky-Golay low-pass filters.

[0058] In the following step (iii) of the procedure of Fig. 8 The filtered ECG signals are then converted into orthogonalized measured values ​​based on vectorcardiography by means of the evaluation device 18, which is connected to the filter 16 for signal processing. This localizes significant regions in the heartbeat signal, e.g., the beginning and end of the QRS complex, the beginning, maximum, and end of the T wave. To determine the actual times of the physiological extreme points, it is advantageous if the aforementioned leads A, D, H, I, and V, which are located between the lead points E1-E4 (cf. Fig. 2a, Fig. 2b ) are each transformed into spherical coordinates.

[0059] Finally, in a further step (iv) of the procedure according to Fig. 8 The orthogonalized measured variables are input into the AI ​​system 20 or a neural network 20 N. It should be noted that the AI ​​system 20 is based on already known diagnostic data from comparison patients, for whom – as already explained at the beginning – there is clear knowledge regarding their state of health ("healthy" or "ill"). Based on this, a diagnosis for the examined patient 11 can then be established within the AI ​​system 20 using the vector data analyzer 115 and the vector data evaluation memory 116 by comparing the input orthogonalized measured variables with the diagnostic data of the comparison patients.

[0060] For step (i) of the procedure of Fig. 8 It is recommended that a non-invasive recording of ECG signals is carried out at the heart 12 of the patient 11 at exactly the four lead points E1-E4, which were described above with reference to the Fig. 2a und Fig. 2b have been explained.

[0061] By means of an advantageous further development or supplementation of the procedure of Fig. 8 it is possible that the AI ​​system 20 or the neural network 20 N is trained before step (iv), namely by inputting specific learning values ​​f. This is shown below using the Fig. 9 shown and explained in detail: The number of specific learning values ​​f with which the AI ​​system 20 or the neural network 20 N is trained before step (iv) can be between 10 and 30, and can, for example, take the value of 20. These specific learning values ​​f are determined by the following sequence of steps: (v) Providing measurements of a set M (cf. Fig. 10 ) of patient 11 with a known finding, these measurements being orthogonalized on the basis of vectorcardiography, (vi) providing a plurality of time series parameters (cf. Fig. 11a and Fig. 11b ) and at least one statistical procedure (cf. Fig. 12 ), (vii) forming a 3D matrix 25 (cf. Fig. 10a ), where the orthogonalized measured variables of the set (M) of patients define the rows, the time series parameters the columns and the time series length the depth of this matrix (25), where in the case of scalar parameters the depth is equal to one, (viii) classifying all value pairs of the 3D matrix (25) according to the principle of the "area under curve" (AUC) calculation, (ix) selecting a value pair from the set according to step (viii) with the highest AUC value, (x) checking a further value pair from the set according to step (viii), and selecting this value pair if a threshold value for a correlation with the value pair from step (ix) is smaller than 1.65 / √N, with N = number of data points orParameter statistics (patients) according to step (vi) (xi) repeating step (x) for a further pair of values ​​from the set according to step (viii), and selecting this pair of values ​​if a threshold value for a correlation with the previously selected pairs of values ​​is smaller than 1.65√N in terms of amount, and (xii) repeating steps (ix) to (xi) until a predetermined number of, for example, 20 pairs of values ​​is reached, which are then defined as specific learning values ​​f and entered into the AI ​​system (20) for training. .

[0062] The above-mentioned steps (v) to (xii) of the further development of the procedure according to Fig. 9 are symbolized in simplified form by blocks in the corresponding flowchart. The arrow "f," which enters the process between steps (iii) and (iv) from below after step (xii), is to be understood as meaning that the predetermined number of specific learning values ​​f defined in step (xii) are input into the AI ​​system 20 or the neural network 20 N.

[0063] With regard to the advantageous further development of the procedure according to Fig. 9 For the sake of explanation, it may be pointed out that in step (v) the measured variables of the set M of patients 11 are provided in the form of time series, preferably in milliseconds or in the form of heartbeats. The formation of the 3D matrix 25 according to step (vii) is symbolically shown in the Fig. 10a The set M of all patients is plotted as the ordinate and can be divided, e.g. from top to bottom, first into a group of healthy patients (e.g. without CHD findings) and then into a group of sick patients (e.g. with CHD findings). The 3D matrix includes the set M of the total of 284 time series, as well as the scalar personal parameters, which are in the Fig. 11a - 11d However, only the 282 time series parameters are used to train the neural network.

[0064] Regarding the 292 parameters included in the Fig. 11a , 11b , 11c and 11d It is specifically noted that these are 284 time series parameters. Eight personal scalar parameters, each relating to a patient under study, are taken into account for probability calculations using Bayes' theorem.

[0065] At this point, it should be specifically noted that in the aforementioned method, in steps (x) and (xi), the individual thresholds for the respective correlations with the value pair from step (ix) do not assume a constant fixed value, but rather depend on the number of heartbeats or the majority of time series parameters according to step (vi). Thus, higher correlations are permitted for short time series (and thus smaller values ​​of N), and vice versa.

[0066] Fig. 10b shows the 3D matrix in a standardized version with a uniform depth, which is achieved by linking or calculating its columns (= patient data) and rows (= time series parameters) with a number of statistical methods, of which six methods are shown as examples in Fig. 12 are shown, namely: Mean variance kurtosis, skewness, 5% quantile, 95% quantile.

[0067] The possible application of these six methods is shown in the illustration of Fig. 10b indicated by the entry of "6" (in the right-hand part of the image). It should be noted that in step (vii), at least one of these statistical methods, or even several such methods, can be applied to define the depth of the 3D matrix 25 and, if necessary, to achieve a uniform depth for the realization of a normalized matrix 25N.

[0068] Furthermore, for the procedure according to Fig. 9 It should be noted that in step (viii), the AUC calculation can be performed empirically or according to the Johnson distribution principle. The Johnson distribution principle is known per se, but with the present invention, it could be used for the first time in connection with the evaluation of patient data or ECG signals for the purpose of early detection of the presence of coronary heart disease and / or HRS.

[0069] As already explained, in order to evaluate the data for the purpose of making a diagnosis, it is advantageous if the AI ​​system 20, into which the specific learning values ​​f are input, has at least one neural network 20 N , or a plurality of such networks 20 N .

[0070] By means of the present invention, it is possible, as explained, on the one hand to carry out a method for the early detection of the presence of CHD and / or HRS of a patient 11 to be examined, as is shown in the flow chart of Fig. 8 On the other hand, in order to improve the diagnosis of the patient 11, it is possible to train the AI ​​system 20 (or a neural network 20 N ) before the actual measurement, as shown in the flow chart of Fig. 9 Such training is carried out on the basis of data from patients whose health status ("healthy" or "ill") is clearly known. In this respect, such training is also referred to as "supervised training" ( supervised learning ) to understand.

[0071] These two possibilities, namely both a preliminary training for the AI ​​system 20 (or a neural network 20 N ) and the actual measurement of a patient 11 for the purpose of establishing a desired diagnosis, are shown again below in the flow charts of the Fig. 13 und Fig. 14 The flow chart of Fig. 13 shows with its blocks 13.1 - 13.5 a sequence of steps for the purpose of training the AI ​​system 20. Such training serves to optimize the actual examination or diagnosis of a patient 11, which is determined by the sequence of steps in the flow chart of Fig. 14 is illustrated.

[0072] In the flow chart of Fig. 13 Step 13.1 essentially corresponds to step (i) of Fig. 8 , wherein step 13.2 essentially corresponds to step (ii) of Fig. 8 The same applies to step 13.3, which essentially corresponds to step (iii) of Fig. 8 To avoid repetition, please refer to the explanations for steps 13.1, 13.2 and 13.3. Fig. 8 be referred to.

[0073] The following step 13.4 in the flow chart of Fig. 13 corresponds to a parameter extraction ( "parameter extraction"), in which - in accordance with steps (v) to (vii) of Fig. 9 - Time series parameters are linked or calculated with the measured variables of a set M of patients 11. Here, too, it proves advantageous if this is done on the basis of spherical coordinates into which the derivatives A, D, H, I, and V have been suitably transformed.

[0074] The subsequent step 13.5 aims at determining the feature ( "feature evaluation") and essentially corresponds to a sequence of steps (viii) to (xii) of Fig. 9 , which have already been mentioned and explained above. This means that with this feature determination according to step 13.5, the specific learning values ​​f are determined or ascertained.

[0075] Following this, the flow chart of Fig. 13 , namely in step IV T , the specific learning values ​​f are input into the AI ​​system 20 (or into a neural network 20 N ), corresponding to step (iv) of Fig. 8 . As a result, the AI ​​system 20 is suitably trained by entering the specific learning values ​​f. At this point, it should be pointed out again that, within the scope of this training, it is of great advantage that the number of specific learning values ​​f is relatively small and can, for example, assume the value 20. Alternatively, the number of specific learning values ​​f can also be less than or greater than 20, and can, for example, be 15 or 25. In any case, with regard to the training of the AI ​​system 20 discussed here using these learning values ​​f, it should be understood that these are always obtained based on the data of a patient 11 for whom clear knowledge regarding their state of health ("healthy" or "ill") is available.

[0076] The flow chart of Fig. 14 The sequence of steps then illustrates the diagnosis actually carried out for a patient 11. Steps 14.1, 14.2 and 14.3 essentially correspond to steps 13.1-13.3, so that in order to avoid repetition, reference can be made to the explanation of steps 13.1-13.3.

[0077] Step 14.5 in the flowchart of Fig. 14 sees a feature selection ( "feature selection") - here, only the data of an examined patient 11 are used based on the recorded ECG signals, which correspond to the specific learning values ​​f previously determined according to step 13.5. A comparison of these selected features or values ​​with the predetermined specific learning values ​​f then takes place in step 14.6 ("trained network"). Following this, in step IV D, the actual diagnosis IV D is made for the patient 11 with the help of the trained AI system 20, corresponding to step (iv) of Fig. 8 .

[0078] The above-mentioned diagnosis of a patient 11, which was made with a facility 10 of Fig. 1 and by a process according to the flow chart of Fig. 8 or Fig. 14 is carried out, as well as the training of an AI system 20 (or a neural network 20N), which is carried out according to the flow chart of Fig. 9 or Fig. 13 are always based on the sensors S1-S4 being attached to the body of the patient 11, e.g. using the T-shirt of Fig. 3 . According to a further method, which is not part of the invention, any incorrect positioning of these sensors S1-S4 on the body of the patient 11 is to be compensated, so that a realistic diagnosis for the patient 11 is still ensured. Such a method, hereinafter referred to as the "correction method", initially provides that a spatial vector 24 related to the heart 12 of the patient 11 (cf. Fig. 7a ), which represents the vector of the electric field formed by the activity of the heart 12, is determined and suitably displayed. Measurement values ​​of the heart 12 are recorded on the patient's body 14 at a total of four recording points E1-E4, as already described above in connection with the Fig. 2 has been explained. In the correction procedure just mentioned, the derivatives A, D, H, I and V (cf. Fig. 5a, Fig. 5b ) an orthogonal system with the relations x = D cos 45 ° − I y = D sin 45 ° + A z = V − H sin 45 ° formed, whereby the measured values ​​for the patient 11 and the resulting space vector 24 are mapped in this orthogonal system x, y, z.

[0079] The correction procedure then looks like this, with reference to the representations in the Fig. 5a, Fig. 5b , Fig. 15 und Fig. 16 In detail, the following steps: (a) Carrying out a measurement on a patient using the first to fourth lead points E1-E4 on the patient's body 14 in order to obtain a cardiogram for this patient, (b) Extracting the amplitudes of the R-wave from the cardiogram from step (a) for each heartbeat in the x-, y- and z-direction, (c) Determining mean values ​​µ x , µ y , µ z and standard deviations σ x , σ y , σ z of the amplitudes recorded in millivolts from the cardiogram according to step (b), wherein a calculation vector 26 is then formed using these mean values ​​µ x , µ y , µ z and standard deviations σ x , σ y , σ z, (d) Forming a coefficient matrix 28 which is based on a principal axis transformation for measurements on Comparison patients with different investment schemes have been obtained, (e) multiplying the calculation vector 26 of step (c) by the coefficient matrix 28 of step (d),to form a result vector 30 with a total of six principal axes PC 1 - PC 6 , (f) extracting the first principal axis PC 1 and the second principal axis PC 2 from the result vector 30 of step (e), to form a reference point PC 1 , PC 2 in the space of the first and second principal axes, (g) determining a Euclidean distance of the reference point PC 1 , PC 2 from a predetermined target point PC 1 fit< , PC 2 fit< , which corresponds to a correct position of the four derivative points E1-E4 on the human body 14, and (h) if the distance of the reference point PC 1 , PC 2 from the predetermined target point PC 1 fit< , PC 2 fit< is greater than a predetermined maximum value: performing an angle correction for a first triangle 31 formed from the first derivative point E1, the third derivative point E3 and the fourth derivative point E4 and for a first triangle 31 formed from the first Second triangle 32 formed by the derivative point E1, the second derivative point E2 and the fourth derivative point E4,so that the Euclidean distance between the reference point PC 1 , PC 2 and the predetermined target point PC 1 fit< , PC 2 fit< is minimized by adapting the orthogonal system x, y, z to the changed geometry. ,

[0080] The above steps (a) to (h) of the correction method are explained in such a way that the amplitudes with which the calculation vector 26 is formed in step (c) are those measured values ​​which were previously determined by the measurement according to step (a) or in the method of Fig. 8 in step (i), in the process of Fig. 13 in its step 13.1 and / or in the process of Fig. 14 in whose step 14.1 are recorded non-invasively with the ECG signals at the heart 12 of the patient 11. The step (e), after which the calculation vector 26 is multiplied by the 6x6 coefficient matrix 28 from step (d) and the result vector 30 results from this, is shown in the Fig. 15 This result vector 30 is in the form of a 6x1 matrix. In step (f), the third to sixth rows of the result vector 30 are calculated as shown in Fig. 15 marked with a strikethrough, deleted or ignored for further calculation, whereby the first principal axis PC 1 and the second principal axis PC 2 are extracted from the result vector 30 to form the reference point PC 1 , PC 2 . Furthermore, the Euclidean distance, which is determined in step (g) of the correction process, is shown in the diagram of Fig. 16 illustrated.

[0081] The diagram of Fig. 16 also shows an exemplary target area, symbolized here by a rectangle 34 with dashed lines. In this regard, it should be noted that before performing step (h), the predetermined target point PC 1 fit< , PC 2 fit< is selected from this target area, which is formed based on standard deviations of a plurality of verified correct measurements of the target point on the human body 14.

[0082] In step (h) of the correction method, the angle correction defined herein can be used to determine adjustment values ​​ε, η, with which the non-rectangular angles α of the first triangle 31 and the non-rectangular angles β of the second triangle 32 can be corrected. This makes it possible to compensate for an incorrect position of the first derivation point E1 on the human body 14, in particular, in order to take non-rectangular triangles into account if necessary. In this way, the adjustment value ε represents an adjusted value which, in the case of correctly applied electrodes, is identical to the angle α of the first triangle 31. The same applies to the adjusted value η, which, in the case of correctly applied electrodes, corresponds to the angle β of the second triangle 32. This relationship is also shown graphically in the representations of Fig. 5a und Fig. 5b shown.

[0083] The above-mentioned adjustment values ​​ε, η, which can be used for the angle correction of step (e), can be determined by minimizing the Euclidean distance between the reference and the target point.

[0084] The coefficient matrix 28, with which the calculation vector 26 is multiplied in step (e), is formed by a principal axis transformation of a 23x6 matrix based on 23 test measurements and the mean values ​​and standard deviations of the measured values ​​determined therefrom.

[0085] The correction method explained above is based on the principle of a principal axis transformation, which can compensate for an incorrect fit of electrodes or sensors S1-S4 on the human body 14. This applies in particular to the position of the sensor S1, which is assigned to the first derivation point E1, and is advantageous, for example, in the case that the measured values ​​of the heart 12 with the T-shirt 22 of Fig. 3 be recorded.

[0086] The above discussion of Fig. 7a It has already been pointed out that the spatial vector 24 shown therein, which is determined using vectorcardiography, represents the electrical activity of the heart 12. For the present invention, a cardiac monitoring method is thus defined in which a quotient is formed from the areas swept over by a length of the spatial vector 24 (= radius vector) as a function of time during the R-wave and during the T-wave, respectively, and this quotient is subsequently subjected to further evaluation. These areas swept over by the spatial vector 24 as a function of time are shown by way of example in the diagram of Fig. 17 shown.

[0087] For the present invention, it has been found that for the examined heart 12 of a patient 11, ischemia or coronary heart disease (CHD) is detected if the quotient formed from the areas swept over by the space vector 24 during the R-wave and during the T-wave Fläche R − Welle Fläche T − Welle = a lies outside an interval between the limit values ​​a 0,KHK and a 1,KHK.

[0088] A special case of the above-mentioned findings is the case where a cardiac arrhythmia (HRS) is detected for the examined heart 12 if the quotient formed from the areas swept over by the space vector 24 during the R-wave and during the T-wave satisfies the condition Fläche R − Welle Fläche T − Welle < a 0 , HRS or the condition Fläche R − Welle Fläche T − Welle > a 1 , HRS fulfilled.

[0089] With regard to the limit values ​​a 0,CHD and a 1,CHD or a 0,HRS and a 1,HRS , with which the quotient formed from the area (R-wave / area (T-wave) is compared or correlated when examining the presence of CHD and / or HRS in the actual examination of a patient, reference may be made to the explanations in the introductory part of the present patent application in order to avoid repetition, according to which these limit values ​​can also be determined or optimized depending on a training set.

[0090] The various findings which are based on the ratio of the areas covered by the space vector 24 as a function of time during the R-wave and during the T-wave for the last-mentioned method according to the present invention can of course also be applied or taken into account in the previously mentioned methods which are described with reference to the flow diagrams according to Fig. 8, Fig. 9 , Fig. 13 or Fig. 14 have been shown and explained. Bezugszeichenliste

[0091] 10Device for the early detection of CHD or HRS 11Patient 12Heart (of patient 11) 14Body (of patient 11) 16Filter 18Evaluation device 20System based on Artificial Intelligence (AI) 20 N Neural network 22T-shirt 24Spatial vector 253D matrix 25 N normalized 3D matrix 26Calculation vector 28Coefficient matrix 30Result vector 31First triangle 32Second triangle 34Target area 13.1-13.5Steps of a procedure (cf. Fig. 13 ) 14.1-14.5Steps of a procedure (cf. Fig. 14) 100Sensor device 102Data recorder 103Signal receiver 104Signal converter 105Signal storage 106Device 107Signal reader 108Vector data generator 109Vector data memory 110Vector data evaluator 111Vector data synchronizer 112Vector data display 113Server 114Vector data memory 115Vector data analyzer 116Vector data evaluation memory AAnterior lead DDorsal lead E1First lead point E2Second lead point E3Third lead point E4Fourth lead point FSpecific learning values ​​(for the neural network 26) HHorizontal lead IInferior lead MSet of patients (with known findings) PC 1 First principal axis PC 2 Second principal axis S1-S4Sensors corresponding to the lead points E1-E4 are assigned Vvertical derivative αNon-rectangular angles of the first triangle 21 (E1-E3-E4) βNon-rectangular angles of the second triangle 22 (E1-E2-E4) ε, ηAdjustment values ​​for the angle correction, for the angles α and β

Claims

1. A heart monitoring method in which ECG signals are recorded at the heart (12) of a patient and on the basis of which a space vector (24) is determined using vectorcardiography, wherein this space vector (24) represents the course of the sum vector of the electrical field of the heart (12) and has a direction corresponding to the field direction and a length corresponding to the potential, characterized in that a quotient is formed from the areas covered by a length of the space vector (24) (= radius vector) as a function of time during the R-wave and during the T-wave, respectively, wherein this quotient is then subjected to further evaluation taking into account at least one predetermined limit value, in order to verify the presence of coronary heart disease (KHK) and / or cardiac arrhythmia for a patient, wherein coronary heart disease (KHK) is detected for the screened heart (12) if the quotient formed from the areas covered by the space vector (24) during the R-wave and during the T-wave, respectively Area R-wave Area T-wave = a is located outside a range defined by a lower limit value a0,KHK and an upper limit value a1,KHK, and that the limit values a0,KHK and a1,KHK are determined as a function of a training set, wherein the lower and upper limit values a0 and a1 are respectively determined by a training set of test person data, comprising the steps: (i) calculating the quotient area (R-wave) / area (T-wave) for all test persons to obtain time series ratios; (ii) calculating a mean value (µ) and an associated standard deviation (σ) from the time series ratios of step (i); and (iii) determining the lower limit a0 taking into account the mean value (µ) and the standard deviation (σ) of step (ii), by the relationship: a 0 = mean value ( μ ) + standard deviation ( σ ) , and / or determining the upper limit a1 taking into account the mean value (µ) and the standard deviation (σ) of step (ii), by the relationship: a 1 = mean value ( μ ) + standard deviation ( σ ) , wherein in step (ii) the mean value and the associated standard deviation are calculated only for time series ratios of healthy patients ("- -").

2. The method according to claim 1, wherein a cardiac arrhythmia (HRS) is detected for the screened heart (12) if the quotient formed by the areas covered by the space vector (24) during the R-wave and during the T-wave, respectively, satisfies the condition of Area R-wave Area T-wave < a 0 , HRS or the condition of Area R-wave Area T-wave > a 1 , HRS 3. The method according to claim 2, wherein the lower limit value a0,KHK and / or the upper limit value a1,KHK are determined as a function of a training set.

4. The method according to any of the preceding claims, wherein the space vector (24) is determined according to a method which serves for determining and displaying a space vector (24) related to the heart (12), which represents the vector of the electrical field formed by the activity of the heart (12), wherein measured values of the heart (12) are recorded on the body (14) at a first lead point (E1), at a second lead point (E2), at a third lead point (E3) and at a fourth lead point (E4), wherein potential differences are measured in the form of an anterior lead (A) between the first lead point (E1) and the fourth lead point (E4), a dorsal lead (D) between the second lead point (E2) and the fourth lead point (E4), a horizontal lead (H) between the third lead point (E3) and the fourth lead point (E4), a vertical lead (V) between the first lead point (E1) and the third lead point (E3), and an inferior lead (I) between the first lead point (E1) and the second lead point (E2), wherein an orthogonal system is formed with the relationships: x = D cos 45 ° − I y = D sin 45 ° + A z = V − H sin 45 ° and the measured values and the space vector (24) determined therefrom are mapped in this orthogonal system (x, y, z), characterized by the following steps: (a) performing a measurement on a patient using the first to fourth lead points (E1-E4) on the body (14) of the patient in order to obtain a cardiogram for this patient, (b) extracting the amplitudes of the R-wave from the cardiogram of step (a) for each heartbeat in the x, y and z direction, (c) determining mean values µx, µy, µz and standard deviations σx, σy, σz of the respective amplitudes recorded in millivolts from the cardiogram in step (b), wherein these mean values and standard deviations then form a calculation vector (26), (d) forming a coefficient matrix (28) which is obtained based on a Principal Component Analysis by using different formats for measurements on reference patients, (e) multiplying the calculation vector (26) of step (c) by the coefficient matrix of step (d) to form a resulting vector (30) with a total of six main axes (PC1 - PC6), wherein the coefficient matrix (28), with which in step (e) the calculation vector (26) is multiplied to form the resulting vector (30), is a 6x6 coefficient matrix, (f) extracting the first main axis (PC1) and the second main axis (PC2) from the resulting vector (30) of step (e) to form a reference point (PC1, PC2) in the space of the first and second main axis, (g) determining an Euclidean distance of the reference point (PC1, PC2) from a predetermined target point (PC1fit, PC2fit) which corresponds to a correct position of the four lead points (E1-E4) on the human body (14), and (h) if the distance of the reference point (PC1, PC2) from the predetermined target point (PC1fit, PC2fit) is greater than a predetermined maximum value: performing an angular correction for a first triangle (31) formed by the first lead point (E1), the third lead point (E3) and the fourth lead point (E4), and for a second triangle (32) formed by the first lead point (E1), the second lead point (E2) and the fourth lead point (E4) so that thereby the Euclidean distance between the reference point (PC1, PC2) and the predetermined target point (PC1fit, PC2fit) is minimized by adapting the orthogonal system (x, y, z) to the changed geometry, wherein in step (h) the angular correction determines adjustment values (ε, η) with which the non-orthogonal angles (α) of the first triangle (31) and the non-orthogonal angles (β) of the second triangle (32) are corrected, so that at least an incorrect position of the first lead point (E1) on the human body (14) is compensated.

5. The method according to claim 4, wherein prior to performing step (h), the predetermined target point (PC1, PC2) is selected from a target area (34), which is formed on the basis of standard deviations of a plurality of verified correct measurements of the target point on the human body (14).

6. The method according to claim 4 or 5, wherein in step (h) for the angular correction the adjustement values are determined by minimizing the Euclidean distance between reference and target values.