WO2021242097A1

WO2021242097A1 - Ecg based method providing acquired cardiac disease detection

Info

Publication number: WO2021242097A1
Application number: PCT/NL2021/050331
Authority: WO
Inventors: VAN Peter Michael DAM; VAN Eelco Mattias DAM
Original assignee: Ecg Excellence B.V.; Peacs Investment B.V.
Priority date: 2020-05-25
Filing date: 2021-05-25
Publication date: 2021-12-02
Also published as: EP4175548A1; EP4178444A1; WO2021242096A1

Abstract

The present invention relates to a method, such as implemented on a computer, providing cardiac disease detection, relative to the heart in a torso while using ECG measurement data from an ECG recording device, the method comprising steps of: - receiving (100) a 3D or anatomic model and/or processing thereof, such as a heart model and/or a torso model comprising the heart model, - receiving (200a) ECG recordal data and/or processing thereof, preferably with corresponding 3D torso information, of an ECG, performed on the torso - performing (200) beat selection, such as selecting a beat or assembling a beat average based on several selected beats, for subsequent processing based on the beat selection, - determining (300) a VCG and/or mTSI based on the ECG recordal data relative to the model, - determining a detection result for a genetic cardiac disease, preferably based on or comprising at least one of cardiac anatomical VCG and/or mTSI features, such as position or direction, preferably at specific times, time segments or positions relative to the ECG recordal data, - display the detection result in relation to the model.

Description

ECG BASED METHOD PROVIDING CARDIAC DISEASE DETECTION

The present invention relates to a method, implemented on a computer, providing cardiac disease detection, relative to the heart in a torso while using ECG measurement data from an ECG re-cording device.

Brugada syndrome (BrS) is an inherited disorder associated with a high risk of sudden unexpected death due to ventricular arrhythmias (VA) in young and otherwise healthy patients with apparently normal hearts. The diagnosis of BrS is achieved in presence of a both spontaneous or drug- induced type-1 Brugada pattern in 12-lead electrocardiogram (ECG), which under complex circumstances reveals coved elevation of the ST interval specifically observed in V1-V2 precordial leads 4.

However, such diagnosis has been very complex as with such BrS patients, the cardiac electrophysiologic substrate is generally localized at the epicardial level of the outflow tract of the right ventricle (RVOT). So far, this localization may only be demonstrated by complex mapping methods, either by body surface mapping requiring the acquisition of 250 electrodes 6, or by invasive epicardial isopotential mapping.

The inventor of the present invention set out to devise improvements relative to such diagnosis, and to simultaneously provide improvements relative to other diagnosis or at least to provide renderings of results that are readily interpretable as to point to deviations of normal heart functioning or heart behavior. To this end, the present invention relates to a method, such as implemented on a computer, providing cardiac disease detection, relative to the heart in a torso while using ECG measurement data from an ECG recording device, the method comprising steps of:

- receiving (100) a 3D or anatomic model and/or pro- cessing thereof, such as a heart model and/or a torso model comprising the heart model,

- receiving (200a) ECG recordal data and/or processing thereof, preferably with corresponding 3D torso information, of an ECG, performed on the torso

- performing (200) beat selection, such as selecting a beat or assembling a beat average based on several selected beats, for subsequent processing based on the beat selection,

- determining (300) a VCG and/or mTSI based on the ECG recordal data relative to the model,

- determining a detection result for an aquired cardiac disease, preferably based on or comprising at least one of cardiac anatomical VCG and/or mTSI features, such as position or direction, preferably at specific times, time segments or positions relative to the ECG recordal data,

- displaying the detection result in relation to the model.

An advantage of such a method according to the present invention is that a detection result may be provided based on ECG recordal data, which may be interpreted without in-depth knowledge of heart diseases or cardiology for the purpose of directing a patient towards a specialism at low cost. ECG is a low threshold measuring procedure that requires little time. As such the present invention provides indications for deviations of normal heart functioning or heart behavior such that when such indications are found, only the subjects showing this indication may be indicated for specific detailed more complex medical diagnosis or treatment. As such, a straightforward and cost efficient selection of persons for which the indication is found need such further diagnosis steps beyond the present invention. This is very advantageous as obtaining ECG re- cordal data is a very straightforward process that takes very little time and may be performed with a relatively little skill set and with relatively simple equipment, at least as opposed to other medical imaging, such as echo, x- ray or 3D imaging, such as MRI or CT.

Furthermore, the interpretation and subsequent directing of a patient towards a specialism at low cost according to the present invention is achieved with high accuracy, which means that a relatively small number of patients need to be directed towards the specialism.

According to a first preferred embodiment, the steps of performing a beat selection comprises steps of determining at least one fiducial of the selected beat or assembled beat average. Such fiducials may be automatically determined from the ECG curve, such as an onset of QRS, or an offset of a QRS, a QRS 90, a J-point30.

According to a further preferred embodiment, the steps of determining a detection result comprises steps of determining properties of the VCG and/or mTSI. Such properties provide information that may advantageously be made visible relative to the model.

Further preferably, the method comprises steps of determining whether the properties of the VCG and/or mTSI features are within a predetermined thresholds or ranges.

Further preferably, the method comprises steps of determining a projection or displayable result of the properties of the VCG and/or mTSI features. Herewith, it is advantageously achieved that such features provide readily discernible information, especially when shown on a display screen of a computing device performing such methods. Prior art ECG interpretation is performed based on wave shapes of specific leads of an ECG, which may be misinterpreted easily, especially when ECG leads are not placed at specifical- ly defined locations to provide comparability to the results. Interpretations require highly skilled persons to perform such interpretations and misinterpretations are also possible with high skilled persons in case of slight variations of such ECG recordals, such as because of leads that are not placed specific enough.

According to a further preferred embodiment, the method comprises steps of performing a differential diagnostic operation between results of at least two ECG recordings that are recorded at different times, such as a most recent ECG recordal and a previous ECG recordal. Such differential analysis provides the possibility of providing a displayed result of differences over time that are readily readable based on the said features.

A further preferred embodiment comprises steps of relating a location of ECG leads relative to the model, preferably by means of measurement data from a 3D imaging device, such as a camera. With such steps, the placement may be controlled in order to improve the reliability of the results of performing this method.

According to a further preferred embodiment, the method comprises steps of determining at least one patient characteristic, such as torso size and or heart size based on a torso size sensor, such as stretch sensor, input.

Method according to one embodiment comprising steps of determining torso size and/or heart size based on patient information input, or from a database.

Method according to an embodiment comprising steps of using a standard model, such as obtainable from a database.

Method according to an embodiment comprising steps of creating at least one altered model, preferably the heart model based on at least one disease based differential, such as changes or defects. Method according to an embodiment comprising steps of determining a VCG and/or mTSI for at least one of the at least one altered model.

Method according to one embodiment comprising steps of determining activation and recovery derived parameters for at least one of the at least one altered model.

Method according to one embodiment comprising steps of determining at least one of cardiac anatomical VCG and/or mTSI features, such as position or direction, preferably at specific times, time segments or positions relative to the ECG recordal data.

Method according to an embodiment comprising steps of selecting a best match of the at least one model for matching boundary conditions of the respective mTSI and/or mTSI feature and the respective disease.

According to a further embodiment, the method comprises steps of estimating the mean path of cardiac activation and recovery using the VCG direction.

According to a further preferred embodiment, the method comprises steps of using the mid QRS, the center of mass or 50% activation as the start position of estimating the mean path of cardiac activation and recovery.

According to a further preferred embodiment, the method comprises steps of plotting the mTSI position per X, Y and/or the Z axis relative to the start of the mTSI.

According to a further preferred embodiment, the method comprises steps of plotting the movement of the mTSI relative to the same of a population of subjects.

A further aspect according to the present invention comprises a system or computer for performing of a method according to one or more of the previous claims.

Further advantages, features and details of the present invention will be further elucidated on the basis of a description of one or more preferred embodiments with reference to the accompanying figures. Similar yet not necessarily identical parts of different preferred embodiments may be indicated with the same reference numerals.

Fig. 1 comprises a flowchart relating to a first preferred embodiment according to the present invention.

Fig. 2-6 comprises flowchart relating to preferred embodiments according to the present invention.

Fig. 7-29 comprises graphical representations to be rendered visibly to a person according to preferred embodiments according to the present invention.

Fig. 30 shows a preferred embodiment of a system according to the present invention.

A first preferred embodiment (Fig. 1) according to the present invention relates to a method providing cardiac disease detection. The flowchart provides a general overview of such an embodiment.

The detection of any cardiac disease from the ECG assists in optimizing patient treatment to a personalized level. Especially changes in the genetical disease patients, like Brugada syndrome, ARVC, ACM, HCM, etc, and distinctly separate acquired cardiac diseases, for instance ischemia or Arterial Coronary Syndrome (ACS), may be monitored reliably with the ECG. The relevant changes are readily detectable with the method according to embodiments as described below and are as such of diagnostic use.

In step 100, an ECG, such as a baseline ECG is provided or received, such as from an ECG providing a live measurement or from an earlier recorded measurement. Optionally, positions of the electrodes on the chest are measured with a 3D camera. Other methods of recording the ECG location are envisaged in relation to edge embodiment, such as by means of identifying the location of the ECG electrodes within a stationary or alternating electrical, magnetic, or electromagnetic field.

In step 200, a vector cardiogram (VCG) is determined based on the recorded ECG at least one selected ECG beat, such as atrial or ventricular. 3D signals may be related to the heart model.

In step 300, a mean temporal spatial isochrone (mTSI) is determined based on the computed VCG. To be able to compute the mTSI the initial position of the mTSI is set to the center of mass of either the atria or ventricles, depending on which waveform is analyzed. A velocity used may be determined on the part of the waveform analyzed.

In step 400, features derived from the mTSI and VCG are determined and preferably related to the cardiac anatomy. In step 500, a determination is made of a detection result, such as preferably relating to a current disease state. In step 600, a differential diagnostic relating to waveform changes over time is performed, such as to enable personalized differential ECG diagnosis. Preferably the results are displayed on a display device to be observed.

According to a further preferred embodiment (Fig. 2), certain determinations as to the model are performed. It is preferred that the method is preceded by obtaining an ECG recordal, such as to connect the wires of the ECG recorder to the electrodes on the chest of the patient. For the standard 12 lead ECG this the left and right arms, the left foot and the precordial leads V1-V6 over the heart as shown in Fig. 15.

In step 110, it is determined whether a 3D camera is available to localize the ECG electrodes on the chest. In step 115, a 3D photo is recorded using a 3D camera. In step 120, fiducial markers on the 3D image, such as disclosed in co-pending application pct/nl2018/000021, that is incorpo- rated here as a reference. In step 125, the 3D photo is analyzed and distinct body parts are identified. In step 130, respective ECG cables and electrodes are identified. Shape and colors are different. To be able to localize the electrodes automatically these electrode/cable connector features need to be retrieved from a database.

In step 135, the ECG electrodes are localized, such as by using the shape and color features of the electrodes and ECG cable used. The features in the photo that match database features are used to find and localize the specific ECG electrodes on the chest

In step 140, patient characteristics are determined, such as chest circumference, chest width, chest length, etc.

In step 150, it is checked whether a thorax size sensor based on the ECG electrode system attached is available.

In step 160, an alternative thorax size sensor is used based on the ECG electrodes attached. One implementation of such a sensor is a special (multi-electrode) ECG electrode patch, with stretch and/or bend sensors to measure the thorax size and the chest circumference available. Another implementation preferably uses an impedance measured between different electrodes of the ECG system.

In step 170, it is checked whether clinical patients characteristics are available. 180: The thorax size is preferably estimated from clinical patients characteristics, such as height, weight, gender, etc. 190: When also no clinical patients characteristics are available a standard model will be used. Step 195 relates to several disease specific parameters for the selected model may be added.

For patients associated with Brugada the cardiac model may be enriched with an epicardial layer with different tissue properties than the healthy myocardium. To test various location of such affected epicardial tissue several models with epicardial "Brugada" tissue may be added to simulate the changed activation or recovery of the patients heart. The model that produces the best match between clinical measurements (e.g. ECG) and the simulation results is then selected in further analysis. Selecting the best model will be done in 400. This block just creates a list of potential models.

Alternatively, for an ACS patient, disease specific parameters for the selected model may be added. Alternatively, for patients associated with acute ischemia or myocarditis, see Fig. 22 for examples with resulting vectors, the cardiac model may be enriched with regions of transmembrane (TMP) altered activity (e.g. rise in resting potential, change in recovery time etc.). To test various locations of this changed tissue properties several models with changed TMP behavior may be added to simulate the changed activation or recovery of the patients heart. The model that produces the best match between clinical measurements (e.g. ECG) and the simulation results is then selected in further analysis. Selecting the best model is preferably performed in step 400, preferably based on a list of potential models created in step 195.

Fig. 2 relates to an embodiment for measuring ECG and selecting a heartbeat. In step 200, the method starts. In step 210, the ECG is measured. In step220, a heartbeat, this may also a signal averaged beat, i.e. the average of multiple beats with similar waveform, is selected. In step 230, beat fiducials are selected. Fiducials of a single ECG beat from the 12 lead CEG are, preferably automatically, derived from the root mean square (RMS) of all ECG signals measured. With this, a QRS onset is preferably defined as the time when subsequent ECG samples have an increasing value for at least 10 ms. The QRS offset may be defined as the time when the RMS amplitude is lowest between 80 and 200 ms after a detected QRS onset. QRS90 is preferably defined as the time 90 ms after QRS onset. A J-Point 30 is preferably defined as the time 30 ms after the QRS offset, and the W-point is preferably defined by the intersection point at the time axis and the upslope tangent between the T-peak and the mid-amplitude T-wave (23). Similarly, the T- wave end is preferably defined by the inter section point at the time axis and the downslope tangent between the T wave peak and the mid T wave amplitude (21, 22 ), refer to Fig. 17.

In step 300, a VCG and mTSI is determined or computed as also indicated in the remainder of this document. Reference by incorporation is also made to pct/nl2018/050228 and pct/nl2018/050229.

Fig. 4 relates to an embodiment for computing or determining cardiac anatomical mTSI/VCG features. The method starts in step 400. The mTSI and VCG may be derived from the ECG. The path of the mTSI, however may take into account that certain parts of the myocardium function differently, or are different in size. For instance patients suffering from Brugada, also have a right chamber that may be significantly enlarged. This is however depending on the diseased state. Also some parts of the myocardium may show altered functionality (see also description of step 195 in several variants). For this purpose the list of models generated or listed in step 195 may be tested in this embodiment to produce an mTSI that best fits the diseased heart. The pathway of the mTSI preferably remains within the myocardial envelope. This depends on the used velocity and the shape of the heart. Both may be determined from these mod- els. For instance the velocity of the mTSI moving towards an area with altered properties may be reduced or increased, depending on the parameters of the model.

In step 410, for the list of models with different disease specific properties the mTSI parameters are created. In step 420, for each model the mTSI and other activation and recovery derived parameters are generated. In step 430, the model that best matches the boundary conditions of the mTSI and disease is selected, preferably Such that the: mTSI remains well within the boundaries of the myocardium either atrial or ventricular mTSI remains away from non-active myocardial tissue or is attracted to regions with prolonged ac tivity (epicardial tissue in RVOT for Brugada, or a region of ischemia in ACS).

Fig. 5 relates to an embodiment for performing diagnostic steps. A preferred embodiment of diagnostic steps is performed with a model as selected in step 400 from the list such as preferably generated in step 195. In step510, properties of the QRS are determined. A preferred QRS is as follows, e.g. A transseptal vector is present, QRS duration is less than 100 ms, the ter is less than 40% and the minimal ter > 10 %. In step 520, the ST segment is normal, e.g. no elevation or depression in several of the ECG leads, or the average amplitude of the ST segment (0-80 ms after QRS end) divided by the peak T-wave amplitude of the RMS signals ( see 230) is preferably less than 0.33, preferably above 0.05, in which (the latter may mean that there is a minimal T-wave which may for instance indicate subendocardial ischemia.

In step 530, an mTSI T wave direction is determined.

A normal T wave is directed towards the apex. When the com- plete T wave mTSI is directed towards or generally between the (left or right) apex the T wave is preferably indicated as normal. In step 540, when all blocks of 510-540 indicate normal the beat meets the normal criteria.

In step 550, the direction of the ST segment is determined, i.e. The early phase of the ST (end QRS - 180 ms. In step 560, the T-wave vector directions are determined.

In step 570, preferably five directions of the complete QRS- T-wave sequence may be defined:

1)initial QRS, first 20 ms

2)end QRS direction (last 60 ms)

3)early ST end QRS till 180 ms or QRS end till QRS + 60 ms

4)initial T wave: end early ST - T peak

5)T peak till T-wave end.

Examples.

Fig. 18 provides an example relating to an LcX ischemic event (ACS)

Fig. 19 provides an example of a COVID patient with localized mTSI directions. The early ST and T-wave directions show a clearly different direction towards the RVOT. The ECG wave forms above and below are very similar. Red (31), Cyan (32), Blue (33) a an important feature of the graphical representations is that the projections of the small circles in the septum indicate the direction of the corresponding mTSI feature. This provides the feature to readily discern a very different position in case of a disease relative to a relatively normal heart. In Fig. 19, the projections are located at the bottom whereas the Covid patient clearly has the projections located substantially at the other side of the septum.

Fig. 20 provides Example of a Brugada patient, notice the yellow circle is indeed located in the RVOT! The T-wave directions are more or less normal.

Here an example of the classification of the RBBB (right bundle branch block classification) likelihood = (120 <= QRSduration <= 190) AND

( 20 <= trandcardiacratio < 100 ) AND

( 0.25 <= terminalQrsMTSI towards RVOT < 1)

AND

( 0.4 <= terminalQrsMTSI towards RV free wall < 1) AND

( 0.1 <= terminalQrsMTSI towards RV BASE <

1)

IF likelihood >= 1 THEN RBBB

Fig. 6 relates to an embodiment for performing steps for Differential analysis. In step 610, it is determined whether historical data is available. In step 620, a previous ECG is retrieved and preferably the ECG electrode positions for this measurement are checked, see Fig. 21. In step 630, mTSI directions and VCG directions are compared or differences determined. Deviations larger than a certain threshold, such as 10%, may be visualized and used for differential diagnosis.

In the remainder of this description, underlying or related explanations of the above method steps are disclosed in further detail.

In spontaneous or Ajmaline-induced BrS patients, epi- cardial electroanatomic voltage maps may detect the arrhythmic substrate (AS), characterized by low-voltage (<1.5 mV) areas, generally located at the upper part of the anterior wall of the RV, which represents the area where malignant VA are originated, reference Chevallier S, Forclaz A, Tenkorang J, et al. New Electrocardiographic Criteria for Discriminating Between Brugada Types 2 and 3 Patterns and Incomplete Right Bundle Branch Block. Journal of the Ameri- can College of Cardiology 2011;58:2290-8 or Serra G, Baranchuk A, Bayes-De-Luna A, et al. New electrocardiographic criteria to differentiate the Type-2 Brugada pattern from electrocardiogram of healthy athletes with r'- wave in leads V1/V2. Europace 2014;16:1639-45.

The AS area is preferably variable for extension and distribution in different BrS patients before and after provocative test with Ajmaline (AJM), and the extent of an AS area is correlated with the risk of VA, see Baranchuk A, Bayes-De-Luna A, et al.. The AS can be ablated by epicardi- al radiofrequency (RF) application, and after AS ablation the inducibility of VA is dramatically reduced; the typical type-1 Brugada pattern is also abolished, and can no longer be induced by Ajmaline infusion.

Previous findings of the inventors showed that in BrS patients, the epicardial AS was highly correlated with the finding of signal averaging late potentials (SAECG-LPs), that could be interpreted as an expression of abnormal epicardial electrical activity. Furthermore, while for many years, BrS was considered a purely electric disease, recent findings of the inventors showed that the typical BrS pattern reflects an extensive RV arrhythmic substrate, even associated with consistent RV mechanical abnormalities, and that substrate ablation abolished both Brugada pattern and mechanical abnormalities.

In many clinical conditions, the QRS duration obtained from 12-lead ECG may be a relevant discriminator between normal and pathological activation sequences. However, while the QRS onset is generally easily detected, the QRS offset is often blurry and therefore difficult to ascertain, particularly in patients with right bundle branch block (RBBB) or in BrS patients. However, the study of these late QRS components is crucial for the detection and discrimination of such conditions. Furthermore, the 12-lead ECG interpretation is indeed a complex pattern recognition, which does not directly associate waveforms with specific cardiac structures.

A new method to measure and localize the direction of the electrical activity occurring during QRS including the early phase of the ST segment. This novel ECG technique utilizes the inverse ECG approach (iECG), and combines ECG data with the cardiac anatomy, obtaining a novel ECG ac^¬ cording to the invention, alternatively named cineECG, with the aim to overcome and facilitate the interpretation of the waveforms of the standard 12-lead ECG.

Exemplary background of ACM (Arrhythmogenic cardiomy^¬ opathy) . Arrhythmogenic cardiomyopathy (ACM) is an inherited disease which is for example characterised by local ac^¬ tivation delay due to myocardial fibrosis and fibrofatty replacement in the right free wall. Activation delay may result in re-entry induced ventricular arrhythmias (VA). Often, patients with ACM have VA or sudden cardiac death as first presentation of the disease. Asymptomatic mutation carriers of a pathogenic mutation for ACM and patients with ACM are frequently followed at dedicated cardio-genetic outpatient clinics. Serial diagnostic tests are performed to detect disease progression. Early recognition of ACM and activation delay is essential to prevent sudden cardiac death. Currently, only the ECG or an invasive electrophysi- ological study (EPS) can be used to asses local activation delay. The ECG according to the prior art can be used as screening tool but lacks the ability to relate the electri^¬ cal signals to the cardiac anatomy.

Exemplary background acute coronary syndrome (ische^¬ mia). Acute myocardial ischemia is at first identified by interpretation of a 12-lead ECG recording mostly performed by emergency services such as ambulance staff. Most of the applied ECG-recording equipment uses integrated analysis software tools to interpret the acquired recordings. It is of utmost importance to recognize if the patient has indeed an acute myocardial infarction, or not. If so and the infarction onset is within certain time intervals the patients can benefit significantly if they are being treated by a percutaneous coronary intervention (PCI) as quickly as possible ¹⁶. Previous research has shown that this can save the patient from myocardial damage and the long-term benefits for the patients are thus huge ¹⁷. However, the interpretation by computer-assisted algorithms of myocardial infarction is still not accurate enough ¹⁸ and therefore the gold standard is that a cardiologist on call decides if the patient needs a direct PCI treatment, pharmaceutical treatment for the moment or does not need immediate treatment by a cardiologist. To speed up this triage process, the teletransmission of the ECG to an expert center is getting more and more common. Still about 20 % of the ACS patients do not get the required acute care due to misinterpretation of the ECG waveforms. There is clear need for computer implemented algorithms to interpret the ECG wavforms.

Recently, the area of ECG analysis has been extended by the so-called ECG-imaging (ECGi). This method makes use of heart models derived from clinical imaging data of the patients to visualize the electrical activation sequence of the heart ¹⁹ _' ²°. Three-dimensional (3D) reconstructions of the heart are attractive as within one eye blink one has access to all data as compared to a 12-lead ECG recording for which it is needed to interpret all 12 signals and need to know a large set of rules. Three-D reconstructions and visualization are within cardiology very popular such as by example for echocardiography and reconstructions based on CT or MRI. It is an aim of the invention to implement the collection of 12-lead ECG recordings fully-automated from ambulance until discharge of the patient, including a PCI treatment procedure. The inventors propose the method of longitudinal ECGi in this setting. Methods VCG/mTSI in Fig. 7, the following is disclosed. Panel a) The torso/heart model used with the 8 of the 9 electrode positions (the VF electrode is not shown). The torso/heart model on the left represents the standard 12 lead ECG configuration, used to analyze the PTB and clinical database ECGs. The torso/heart model on the right shows the model, with the adapted Brugada lead system. Panel b) fiducials of a single ECG beat from the 12 lead CEG are automatically derived from the root mean square (RMS) of all ECG signals measured. The QRS onset is defined as the time when subsequent ECG samples have an increasing value for at least 10 ms. The QRS offset is defined as the time when the RMS amplitude is lowest between 80 and 200 ms after the detected QRS onset. QRS90 is defined as the time 90 ms after QRS onset. J-Point 30 is defined as the time 30 ms after the QRS offset, and the W-point is defined by the intersection point at the time axis and the upslope tangent between the T-peak and the mid- amplitude T-wave (oranges lines). Similarly, the T-wave end is defined by the inter section point at the time axis and the downslope tangent between the T wave peak and the mid T wave amplitude (blue lines)

Mean temporo-spatial isochrone (mTSI). The mTSI, which represents the mean trajectory of the cardiac activation pathway, was defined to displace through the heart with a constant velocity of 0.7 meter per second (m/s) in the 3D direction indicated by the

The velocity of 0.7 m/s is in the physiological range of the myocardial propagation velocity. In greater detail: The VCG, the di- rection of activation, is computed from the 9 electrodes, building the 12-lead ECG by the following equation:

Where

• |r._j — mTSlfis the normalized vector between the mTSI 3D- position in the heart and the electrode position on the thorax (y_eS).

• bV_§·_j(t) is the value at of the ECG at an electrode at time-samp1e t.

• Factor a ,was set to 0 for the x direction and 2 for the y and z directions for the unaugmented extremity leads (VR, VL, and VF).

• For all other leads the «^factor was set to 1 for all directions.

• This formula takes into account and corrects for the two different electrode placement systems utilized for Brs patients and control ECG tracings.

As the mTSI has to stay within the heart geometry can be adapted to ensure this behavior.

In this study, all patients are in normal sinus rhythm, then the cardiac activation starts in the left septum, close to the center of mass of the ventricles ¹⁹ _' ²⁵.

Thus, the center of mass of the ventricular model is used as the of starting point of the niT 51(G).

In Fig. 8, the iECG work-flow in a normal subject is exemplary disclosed. A first step is to convert the 12-lead ECG into a VCG, positioned at the center of ventricular mass, from which the mean temporal spatial isochrone (mTSI) trajectory can be constructed (lower panel on the right). The ventricles are projected in three standard orientations: 1) 4 chambers, Right Anterior Oblique (RAO) and Left Anterior Oblique views (LAO).

The right ventricle is indicated in transparent blue. The blue arrows indicate the Right-to-Left axis (LR-axis), the green arrow the Posterior-to-Anterior axis (PA-axis), and the red arrow the Base-to-Apex axis (BA-axis).

The colors of the VCG and mTSI indicate the time from the QRS onset to the QRS offset (color bar on the right). The trans-septal vector is indicated in red. In this case, the trans-cardiac ratio (TOR) is 11%, as the starting point (red) and the end point (purple) are very close. In this case, the mTSI is located for 55% of the QRS duration in the septal region, while for the 45% of the QRS duration in the left ventricle.

Parameter descriptions

In order to visualize the tempo-spatial localization of the electrical activity pathway, we introduced the new concept of cineECG, representing the moving trajectory of the mTSI within the cardiac anatomic structures. To establish and visualize a quantifiable relation between the cardiac anatomy and the mTSI trajectory, three standard X-ray views on the heart were created from the heart model: a standard 4-chamber view, and right and left anterior oblique views (RAO and LAO, Figure 2). Therefore, the terminal direction of the mTSI can be related to specific structures of the heart, like septum, and RV or LV free walls, or RVOT. An example of the construction and visualization of the VCG and mTSI for a normal activation is shown in Figure 9. In a normal subject, the VCG is mainly pointing towards the LV free wall, with a small initial trans- septal vector. In the mTSI trajectory, the trans-septal vector is clearly visible. Moreover, the mTSI stays close to the septum ending in the mid-base LV area.

Several of the quantitative parameters derived from mTSI or VCG:

• The anatomical VCG direction: The mean direction of the VCG for the different segments of the ECG beat, i.e. P- wave, PQ interval, QRS complex, ST segment (e.g. initial, S-omega point) and the T-Wave. This direction is related to anatomical structures of the used heart model.

• Entry - exit point VCG axis: The point where the VCG axis of a given ECG beat segments enters and leave the heart can be computed. The entry point can give an indication for the ectopic activation origin, the exit region can identify the latest activated region.

• Trans-cardiac ratio (TCR): The trans-cardiac ratio (TCR) is defined as the 3D-distance between the starting and the ending points of the mTSI, coincident with the QRS onset to the QRS offset. As this measure is potentially influenced by the size of the heart, it is weighted by the size of the heart model, resulting as a relative number. From our previous experience, normal cardiac activation is usually associated with a trans-cardiac ratio generally well below 40% 11. As an example, the trans-cardiac ratio for the normal subject shown in Figure 2 was 11%.

• Minimal Trans-cardiac ratio (minTCR): The smallest distance between start of the mTSI and a point on the mTSI path after 50% of the mTSI. This is used to determine a loop pattern in the mTSI path.

• mTSI Spatial Location: At each time sample of the mTSI, its 3D-spatial localization in the heart is determined, and three cardiac areas were defined, either septal, left ventricle (LV), or right ventricle (RV). The ini- tial spatial location of the mTSI is left septal, and it moves according to the direction determined by the VCG. In case of normal activation, the mTSI initially moves trans-septal and then towards the LV, staying close to the septum (Figure 2). The spatial location is computed as the percentage of QRS duration spent by the mTSI in each of the three cardiac areas. As an example, in a normal subject, the mTSI is located for 55% of the QRS duration in the septum, for 45% in the LV, but never in the RV (Figure 2). Similarly the mTSI spatial location can also be determined for the other ECG beat segments (e.g. P-wave and T-wave or even the U-wave)

• Terminal mTSI direction: In order to measure the direction of the electrical activity occurring during the terminal phase of the QRS, or concealed within the ST segment, we computed the location of the terminal mTSI direction at the three time markers related to the end of QRS electrical activity as described above, i.e. QRS90, J-point30, and D-point (Figure 1, panel b). The terminal mTSI direction is quantified as a relative number between -1 and +1, indicating the congruence with the direction of any the three cardiac axes (Figure 2).

• Other mTSI direction: can also be determined for the other ECG beat segments (e.g. P-wave and T-wave or even the U-wave), or any fixed interval relative to P-wave or QRS onset.

• Ventricular transseptal vector: The transseptal vector (either by VCG or mTSI) is detected when at least 10 ms the VCG/mTSI is moving through the septal wall relative to the starting point at the left septal center of mass.

• Initial P-wave vector: The direction of the first 30 -40 ms of the P-Wave should indicate left-basal to justify a sinus node activation. mTSI directions are computed as:

VCG directions are computed as:

Classification Algorithm

Based on the P-wave and /or QRS duration and on the above described mTSI parameters, a classification algorithm can be developed to test the ability to discriminate vari^¬ ous pathologies, such as BrS or ACM tracings, various bun^¬ dle branch block tracings, and ACS tracings. Below is a number of diagnostic classes defined. For each class, a set of parameter values can be used to compute the probability of a certain classification. The probability for a certain parameter is set to either 0 or 1, with 1 indicating the classification met, and 0 the classification not met. The classification was determined by the highest probability score. The following parameter values were used:

1.Normal, QRS duration < 110 ms, TCR 5-38%, terminal mTSI direction towards the LV basal area.

2.Acquired zijn de BBB's

3.IRBBB QRS duration 95-120 ms, minTCR <8%, terminal mTSI towards RV basal area

4.RBBB, QRS duration 120-190 ms, TCR >50%, terminal mTSI towards RV basal area

5.ILBBB: QRS duration 120-150 ms, TCR >30%, terminal mTSI towards LV free wall / basal area

6.LBBB: QRS duration 120-190 ms, TCR >50%, terminal mTSI towards LV

7.ACM: QRS duration 95- 200 ms, TCR >50%, terminal mTSI towards the RV free wall and not to the RVOT or RV base, transseptal vector present, T wave direction towards the LV

8. ACS: QRS duration in normal range, (except when located in areas that are associated with blood supply of the His-Purkinje system (basal right aortic region), ST segment and T wave point in the same direction.

9. Brugada: QRS duration 95- 200 ms, TCR >50%, terminal mTSI towards the RV free wall or RVOT (NOT towards the RV or LV basal area), transseptal vector present.

10. Undetermined: Did not match any of the criteria above.

Brugada examples

In order to visualize the tempo-spatial localization of the electrical activity pathway, we introduced the new concept of "cineECG", representing the moving trajectory of the mTSI within the cardiac anatomic structures. To establish and visualize a quantifiable relation between the cardiac anatomy and the mTSI trajectory, three standard X-ray views on the heart were created from the heart model: a standard 4-chamber view, and right and left anterior oblique views (RAO and LAO, Figure 2). Therefore, the terminal direction of the mTSI can be related to specific structures of the heart, like septum, and RV or LV free walls, or RVOT. An example of the construction and visualization of the VCG and mTSI for a normal activation is shown in Figure 2 and web-movie 1. In a normal subject, the FCG is mainly pointing towards the LV free wall, with a small initial trans-septal vector. In the mTSI trajectory, the trans-septal vector is clearly visible. Moreover, the mTSI stays close to the septum ending in the mid-base LV area.

Electrophysiological epicardial mapping. All Brugada patients underwent a combined endo-epicardial mapping procedure using a three-dimensional (3D) mapping system (CARTO 3, Biosense Webster, CA, USA), as previously described ^1/8 Further details are provided in the Supplementary material online. All maps were obtained at baseline conditions and after drug challenge (Ajmaline up to lmg/kg in 5min). Total signal duration was measured for each potential before and after drug challenge as previously described ⁷ _' ⁸. Measurements were interpreted and validated online by two expert electrophysiologists using CART03 system electronic callipers. The potential duration map (PDM) was created by collecting the duration of each EGM. As a result, a colour- coded map was obtained showing the regions displaying the shortest (red colour) and the longest (purple colour) durations. The electrical substrate area was defined as an area where abnormal electrograms (EGMs) were identified if they met at least one of the following characteristics: (i) a wide duration (>110ms) with fragmented component (>3 distinct peaks); (ii) late component of low voltage amplitude ranging from 0.05 to 1.5mV; (iii) distinct and delayed component exceeding the end of the QRS complex ⁷ _' ⁸.

Results

Results are shown for all study groups for the cineECG parameters, specifically the mean QRS duration, the mTSI spatial location, the transcardiac ratio, and the location of the terminal mTSI direction (from QRS90 to QRS offset).

Control ECG tracings

For the 47 normal tracings, the QRS duration was on average 91 ± 7 ms. The mTSI direction was generally pointing towards the LV basal area, and the mTSI trajectory was compact, resulting in a TCR with an average value of 24% (Figure 3, panel a). The trajectory of the mTSI across the heart was illustrated by the cineECG, showing an initial transeptal vector, then moving to the left chamber (web ma- terial Movie 1). Only three out of the 47 normal tracing (6%) had a trans-cardiac ratio of more than 40%, which is considered large for an activation initiated by the His- Purkinje system (Figure 2), partially overlapping the features of the RBBB and Brugada ECG patterns, although their QRS were less than 110 msec, and their terminal mTSI location was still in the left chamber.

For the 18 RBBB tracings, the QRS duration was on average 147 ± 19 ms. The cineECG showed an initial transeptal vector, then moving towards the RV basal area, representing an activation going from the right apical region towards the right base of the heart (web material movie 2). The TCR value was on average 42%, and the cineECG generally showed an open loop configuration (Figure 3, panel b).

Brugada ECG tracings

In the nine patients with spontaneous type-1 Brugada pattern, the QRS duration was on average 163 ± 28 ms, and the TCR was 51+17%, with an open loop configuration. In the 13 Ajmaline-induced BrS patients, the QRS duration was 152j^25 ms, and the TCR was 40%, always with an open loop configuration. No significant differences between spontaneous or Ajmaline-induced Brugada patients in most cineECG parameters were observed, except the TCR and mTSI spatial location for septum and RV.

Noteworthy, at a difference with normal and RBBB tracings, both in spontaneous and Ajmaline-induced BrS patients, the terminal mTSI direction was mainly and homogenously directed towards the RVOT.

The RVOT localization of the mTSI terminal activation detected by cineECG was congruent with the area of ar- rhythmogenic substrate detected in spontaneous BrS patients by the PDMs simultaneously obtained (see Figure 4, panel a). Similarly, the RVOT localization was congruent with the area of electrical substrate detected by the epicardial mapping after Ajmaline infusion (Figure 4, panel b).

Fig. 12 shows the mean spatial direction of the terminal mTSI. In the 4 chamber and RAO projections, there is no overlapping between normal subjects, RBBB patients, or Brugada patients, with spontaneous and Ajmaline-induced BrS patients having similar behavior, in Ajmaline-induced BrS patients, in LAO projection the terminal mTSI direction is even more directed towards the RVOT than spontaneous BrS patients .

Fig. 19 shows four typical examples of the mTSI trajectory for a normal control (panel a), a patient with a complete Right Bundle Branch Block (RBBB, panel b), a patient with spontaneous Brugada pattern (panel c), and a patient with Ajmaline-induced Brugada pattern (panel d). For each patient, the mTSI is shown by the color line during the QRS, while the terminal mTSI is shown as a solid grey line, indicating the terminal direction of the meanTSI.

Fig. 20 shows the epicardial PDMs of one spontaneous BrS patient (panel a) and one Ajmaline-induced BrS patient (panel b). PDMs and 12-lead ECG tracings were obtained simultaneously. The arrhythmogenic substrate (AS) area is indicated in purple. In both cases, the location of the AS area detected by the PDM and the location of the terminal mTSI, by cineECG was coincident, in both cases located in the RVOT area.

Fig. 11 shows the average terminal mTSI directions for all study groups, 47 normal controls (blue), 18 RBBB patients (green), 9 spontaneous BrS patients (red), and 13 Ajmaline-induced BrS patients (purple), in 4 chamber, RAO and LAO views. The solid lines represent the average values, the dotted lines the standard deviation per study group. The direction of the terminal mTSI is clearly diver- gent in normal, in RBBB and in BrS patients, especially in 4-camber and in RAO view. The terminal mTSI direction is coincident in spontaneous and Ajmaline-induced BrS patients, in several projections except LAO (where is even more directed to RVOT in Ajmaline-induced patients).

Figure 12: Example of an ACM patient. The mTSI moves towards the right ventricular free wall, but not towards the RVOT or basel. The T-wave (G, grey line) is directed towards the LV.

Example ACS: ischemia

Below is a patient with an sub-endocardial apical ischemic area. The amount of ischemia varies over time, as the 3 recording taken at different moments in time show.

The electrodes were repositioned at the exact same location using the 3D camera and a projector to indicate the exact location on the chest wall.

Fig. 13 shows three ECGs recorded measured a 2-week interval. The blue one is the first one, red the second one, and black the last one. The ischemia is visible on lead V2 and V5 (lower ST segment). As can be appreciated from these sequential recordings is that the amount of ischemia varies over time. The black ECG shows no ST change in V5. T wave morphology changes are related to heart rate changes.

Fig. 14 shows The mTSI of the last recording shown in Figure 6. Although the ECG shows no ischemia anymore, the ST and T wave direction are pointing towards the right, which is congruent with the small ischemic are in the left apical region. The T-wave direction (both in VCG and mTSI ) related to the cardiac anatomy thus improve the detection and localization of (acute) ischemic areas.

As an alternative to position the VCG at the center of ventricular mass, and embodiment provides placing the origin at the electrical middle. Preferably this means placing the origin at the point of 50% activation or at the middle of the QRS. An advantage hereof is that from this point, all of our timings of the activation of the heart may be derived, which leads to the start of the activation or PVC. The ECG according to the present invention thus provides the opportunity to calculate the location at the start, such as at 0 ms, 1ms and so on. With this, it is possible to determine the location of the start of the activation preferably based on each ECG measurement.

Advantages of ECG according to embodiments include a feasible and pragmatic solution for ECG waveform interpretation using non-invasive cardiac modeling, relating the ECG to the cardiac anatomy, standardizing the ECG visualization, and showing an increased diagnostic value for the ECG. Furthermore, it provides a pragmatic way to directly relate the ECG in a standardized way to the cardiac anatomy. Also, it may increase the clinical and diagnostic value of the ECG in: conduction disorders, ST and T-wave deviations in an acute setting, as well as support Brugada patterns even with low amplitudes in the ST segment. Also, it is easy to implement in existing clinical technology and workflow. Also envisaged is correction for ECG electrode positions and body build using a 3D camera to increase diagnostic value even further.

The interpretation of electrocardiogram (ECG) waveforms according to the prior art comprise deficiencies for all physicians at all training levels, even after extra ECG reading education. The inability to interpret the ECG with adequate accuracy in clinical practice hampers effective ECG usage. Even the extra training of staff does not result in enough improvement of the ECG diagnosis. One of the in- terpretation problems identified by the present inventors is caused by the interindividual variability in the normal ECG waveforms, such as influenced by gender, electrode positions during the ECG recording, and specific electrophys- iological properties of the atria and ventricles. Another problem with the ECG interpretation may attributed to the differences in amplitude, for instance the first 20 ms of the QRS are difficult to interpret without detailed expert knowledge. The present invention and embodiments clarify parts of the ECG with low amplitudes such as by using the direction of the cardiac electrical activity to estimate the mean trajectory of cardiac activation and recovery.

Herewith, the mean tempo-spatial isochrones (mTSI) trajectory according to the invention or embodiments, i.e. the mTSI position on the major cardiac axes: X) posterior- anterior, Y) left-right, and Z) base-apex. The goal of these 3 mTSI graphs is to support the detection of deviating waveforms of QRS and T-wave, even by limited trained clinical personnel. With application of the invention, it is easily performed by representation of the graph relative to graphs of a number of representatives of a population, see Figs. 22-29 as disclosed in greater detail in this description.

The normal ventricular activation is initiated by the Purkinje system. This specialized ventricular conduction tissue rapidly distributes the electrical activation through the heart [8, 9]. The dense and widely branched endocardial system of Purkinje fibers originates from the His-node from which several major branches innervate the left and right ventricle. The electrocardiographic (ECG) signals are the result of multiple activation waves initiated from the different endocardial positions of the left and right ventricle. Consequently the ECG waveforms of the activation waves, the QRS complex , dependent on the timing and anatomical positions of the early break-throughs. The resulting normal mean QRS axis, however, shows limited correlation with the anatomical axis of the heart, indicating the initiation sites are not related to the cardiac anatomy [12]. The ventricular repolarization process, depicted as the T-wave following the QRS, is a much slower process than the activation process, and consequently the local electrical gradients within the myocardium are smaller. The direction of both the overall depolarization and repolarization gradients are used to compute the mTSI trajectory through a standard ventricular geometry.

One goal of this invention is to provide the ability of the ECG to represent the normal ECG relative to a standard cardiac anatomy, ultimately to improve the diagnosis of ECG waveforms.

Methods

CineECG method

The mean temporal spatial isochrone (mTSI) pathway describes the average of all cardiac activity during the activation and recovery of the heart. In short, see Fig.

22:

A) The input to CineECG is the standard 12 lead ECG,

B)The 12 lead ECG is converted into the vectorcardiogram (VCG) using the model of a thorax and heart of e.g. a 57 year normal male with standard positions of the 12 lead ECG electrodes on the chest. The computed VCG is a 3D representation of the direction of cardiac activity per time step.

C)In the next step mean temporo-spatial isochrone (mTSI) trajectory is computed, i.e. the position of the average electrical activity of the ventricles. To position the mTSI trajectory within the cardiac anato- my the mid QRS position of the mTSI is set to the cen^¬ ter of mass of the used ventricular model. This posi^¬ tion represents the center of electrical activity, when approximately half of the ventricular mass is ac^¬ tivated. From this position the mTSI position is computed according:

in which r is the velocity with which the mTSI travels through the myocardial anatomy. For the QRS the veloc^¬ ity is set to 0.7 ms^-1, and for the STT segment the velocity is set to ensure the length

of QRS and STT segment have the same length. The used velocity of 0.7 m/s is chosen to be in the physiologi^¬ cal range of the myocardial propagation velocity.

D) In a final step the mTSI position per X, Y, and Z axis is plotted relative to the start of the mTSI (

E) Fig. 22D), where the X-axis is the posterior-to- anterior axis, Y-axis the right-to-left axis, and

Z(axis the base-to-apex axis. Plotting the movement of the mTSI position relative to the initial position en^¬ ables the comparison between ECGs of different sub- jects.

Fig. 22 shows The workflow of the computation of the CineECG. A) the CineECG input is the standard 12 lead ECG assuming standard 12 lead ECG electrode positions.

B) The ECG is converted into the vectorcardiogram, rep^¬ resenting the direction of cardiac activity through the heart beat. Notice the differences in amplitude between QRS (in white-red) and ST-T-wave (yellow-blue). C) The VCG direction is used to estimate the mean temporo- spatial isochrone position for the QRST sequence. D) The C,U,Z components pf the mTSI are plotted relative to the first mTSI position in the heart. For this normal example the mTSI position moves initially trans-septal, i.e. to the right, subsequently back to the left as the LV has more mass, and finally the T-wave mTSI position moves toward the apex

ECG data

As reference for normal activations, we utilized all 65xx ECGs labelled as healthy control in the certified Physionet PTB XL Diagnostic ECG Database. For each of these ECGs the mTSI and derived parameters were computed for computed median beats.

Results are as follows. The distribution of mTSI positions along the XYZ axes is shown in Fig. 23. The 90% distribution range of the normal mTSI positions along the 3 heart axis (orange lines) in combination with the mTSI po^¬ sitions as shown in Figure 1 (dark green line). All samples of the mTSI wall within the normal distribution.

A direct relation between ECG and cardiac anatomy may improve the characterization and the diagnostic accuracy in a real clinical condition.

The electrotonic interaction, due to the cell-to-cell coupling, plays a significant role in the recovery process, i.e. areas that are activated early recover late and vice versa.

Simulations have shown that small differences in local repolarization processes of the myocardium already result in large T-waveform changes [21], indicating the sensitivity of the T-waveform to local differences in the repolarization duration.

These repolarization changes are so be reflected in the mean path of recovery as represented by the mTSI. For many decades the vectorcardiogram was used to study the spatial direction of cardiac activity, either activation or recovery. However, the relation between amplitude and direction of the vector and cardiac anatomy is complex, and thus difficult to teach and learn. Moreover the vectorcardiogram is also sensitive to electrode positions. All these factors make it difficult to discover intra-QRS conduction delays from just the electro- and vectorcardiogram. The mean temporal spatial isochrones method (mTSI), see elsewhere referenced patent applications, relates the ECG directly to a activation and recovery path within the heart. This method can be used to detect intra-QRS waveform changes related to conduction disturbances.

The repolarization process, depicted as the T-wave following the QRS, is a much slower process than the activation process, and consequently the local electrical gradients within the myocardium are smaller. Consequently the electrotonic interaction, due to the cell-to-cell coupling, plays a significant role in the recovery process, i.e. areas that are activated early recover late and vice versa.

Simulations have shown that small differences in local repolarization processes of the myocardium already result in large T-waveform changes, indicating the sensitivity of the T-waveform to local differences in the repolarization duration. These repolarization changes are also reflected in the mean path of recovery as represented by the mTSI.

Fig. 23 shows the graphs the cardiac axis view of Fig. 22 relative to a population of subjects. It is clearly shown that the more emphasized graphs of the subject under study falls well within the extremes of this population, albeit somewhat near the outside of the population in at least the anterior view.

Fig. 24 in addition shows a representation of the mean temporo-spatial isochrone (mTSI) of the subject below the graphs. It is clearly shown that the mTSI falls well within the confines of the heart in these views.

In Fig. 25, the situation of an LBBB arises in the subject. The graphs clearly show how the person falls well outside the extremes of this population, thus indicating that something is amiss. Also, in the mTSI representations, it clearly shows that the endings fall well outside of the confines of the heart. With this, representations according to such embodiments clearly show an easily interpretable misfunctioning of the heart.

In Fig. 26, a Brugada is present. It is also very clear from the graphs that these fall well outside of the scope of the population. Also the trajectory of the mTSI shows readily interpretable deviations from a normal, such as that of Fig. 24.

A further ailment that can be detected readily from the graphical representations is shown in Fig. 29 in which the person suffers from COVID 19.

Fig. 30 shows An embodiment of a system according to the invention comprises a system for performing a computer implemented method. A computer 5 comprises a processing unit, a with the processing unit functionally coupled memory, a 3D or anatomic model receiver, the model preferably being a torso model and/or a heart model, a location information receiver of at least one heart conduction feature, such as the His Purkinje system or parts, or a model thereof, the base of the papillary muscles, the right free wall moderator band connection, the right apical septal position, the mute apical left septal position and/or the basal left septal position. The computer is preferably cou- pled with an ECG receiver for receiving (600) ECG recording data, preferably with corresponding 3D torso information, of a currently performed ECG. It is however also envisaged to perform the method at a later time based on previously acquired data, including the 3D model, ECG measurements, for creating an activation map and thereto related renderings or data creations as indicated in this document. The computer comprises a heart activation map determining module relative to the model, such as comprising steps of updating of electrophysiological properties of the model. The system may comprise a three-dimensional camera 2, for detecting ECG electrodes arranged at a torso T, is arranged above the torso T (schematically shown) of a person. The camera is suitable for moving thereof relative to the torso such that from several sides the torso can be recorded for detecting of the ECG electrodes jet in place. Data from the camera are transferred to the computer 5. The computer is connected to a monitor 7, keyboard 8 and mouse 9 for receiving input data from these peripherals from a user and for outputting of image data to the user. The computer is furthermore coupled with an ECG amplifier 6 that in its turn is coupled to ECG electrodes 3 on the torso T. A practical number of electrodes that is supplied is between 4 and 16, preferably substantially 12. A larger number for achieving a higher resolution is envisaged and use thereof dependent on the surroundings in which the installation is applied also usable. The skilled person would be able to determine the number of electrodes as a correct choice based on available equipment.

The computer comprises a set of candidate positions determining module, preferably comprising a target position determining module. The system comprises an ECG device 2 for recording of respective ECG sessions, preferably a standardized ECG session, further preferably a standardized 12 lead ECG session. The computer comprises a location information receiver, preferably as a software module, for receiving (500) location information of at least one cardiac vein, such as by estimation or scanning data. The computer comprises a potential implant areas determining module, for the purpose of determining the areas for implantation of at least one pacing system lead, such as for the purpose of outputting such information relating to the model and the potential implant areas for displaying thereof.

Clauses

1. Method, such as implemented on a computer, providing cardiac disease detection, relative to the heart in a torso while using ECG measurement data from an ECG recording device, the method comprising steps of:

- receiving (100) a 3D or anatomic model and/or processing thereof, such as a heart model and/or a torso model comprising the heart model,

- determining a VCG and/or mTSI based on the ECG recordal data relative to the model,

- determining a detection result for an acquired cardiac disease or syndrome, preferably based on or comprising at least one of cardiac anatomical VCG and/or mTSI features, such as position or direction, preferably at specific times or positions relative to the ECG recordal data, - display the detection result in relation to the model. Method according to clause 1 comprising any feature according to the below claims and/or above description.

The present invention is described in the foregoing on the basis of preferred embodiments. Different aspects of different embodiments are expressly considered disclosed in combination with each other and in all combinations that on the basis of this document, when read by a skilled person of the area of skill, fall within the scope of the inven- tion or are deemed to be read with the disclosure of this document. These preferred embodiments are not limitative for the scope of protection of this document. The rights sought are defined in the appended claims. kk kkk

Claims

- receiving (200a) ECG recordal data and/or processing thereof, preferably with corresponding 3D torso information, of an ECG, performed on the torso,

- determining a detection result for an acquired cardiac disease, preferably based on or comprising at least one of cardiac anatomical VCG and/or mTSI features, such as position or direction, preferably at specific times, time segments or positions relative to the ECG recordal data,

- displaying the detection result in relation to the model.

2. Method according to claim 1, wherein the steps of performing a beat selection comprises steps of determining at least one fiducial of the selected beat or assembled beat average.

3. Method according to claim 1 or 2, wherein the steps of determining a detection result comprises steps of determining properties of the VCG and/or mTSI.

4. Method according to claim 3 further comprising steps of determining whether the properties of the VCG and/or mTSI features are within predetermined thresholds or ranges.

5. Method according to claim 3 or 4 comprising steps of determining a projection or displayable result of the properties of the VCG and/or mTSI features.

6. Method according to one or more of the preceding claims comprising steps of performing a differential diag^¬ nostic operation (600) between results of at least two ECG recordings that are recorded at different times, such as a most recent ECG recordal and a previous ECG recordal.

7. Method according to one or more of the preceding claims comprising steps of relating a location of ECG leads relative to the model, preferably by means of measurement data from a 3D imaging device, such as a camera.

8. Method according to one or more of the preceding claims comprising steps of determining at least one patient characteristic, such as torso size and or heart size based on a torso size sensor, such as stretch sensor, input.

9. Method according to one or more of the preceding claims comprising steps of determining torso size and/or heart size based on patient information input or from a da^¬ tabase.

10. Method according to one or more of the preceding claims comprising steps of using a standard model, such as obtainable from a database.

11. Method according to one or more of the preceding claims comprising steps of creating at least one altered model, preferably the heart model based on at least one disease based differential, such as changes or defects.

12. Method according to claim 11 comprising steps of determining a VCG and/or mTSI for at least one of the at least one altered model.

13. Method according to claim 11 or 12 comprising steps of determining activation and/or recovery derived parameters for at least one of the at least one altered model.

14. Method according to one or more of the claims Ills comprising steps of determining at least one of cardiac anatomical VCG and/or mTSI features, such as position or direction, preferably at specific times, time segments or positions relative to the ECG recordal data.

15. Method according to one or more of the claims 11- 14 comprising steps of selecting a best match of the at least one model for matching boundary conditions of the respective mTSI and/or mTSI feature and the respective disease.

16. Method according to one or more of the previous claims comprising steps of estimating the mean path of cardiac activation and recovery using the VCG direction.

17. Method according to one or more of the previous claims using the mid QRS, the center of mass or 50% activation as the start position of estimating the mean path of cardiac activation and recovery.

18. Method according to one or more of the previous claims comprising steps of plotting the mTSI position per X, Y and/or the Z axis relative to the start of the mTSI.

19. Method according to one or more of the previous claims comprising steps of plotting the movement of the mTSI relative to the same of a population of subjects.

20. System or computer for performing of a method ac- cording to one or more of the previous claims. kk kkk