WO2008003828A1

WO2008003828A1 - Method and arrangement for detection of acute myocardial ischemia

Info

Publication number: WO2008003828A1
Application number: PCT/FI2007/050409
Authority: WO
Inventors: Helena HÄNNINEN; Lauri Toivonen; Paula Vesterinen; Heikki VÄÄNÄNEN
Original assignee: Licentia Oy
Priority date: 2006-07-05
Filing date: 2007-07-04
Publication date: 2008-01-10

Abstract

The present invention relates to a method for detecting myocardial ischemia in a person, which comprises the steps of recording ECG-data from the person; transferring the ECG-data in digital form into a computer based automatic analysing system; calculating a QT integral from the ECG-data by using the computer based automatic analysing system; and receiving an interpretation of the QT integral values for detecting persons having myocardial ischemia.

Description

Method and arrangement for detection of acute myocardial ischemia

This invention relates to a new method and arrangement for detecting acute myocardial ischemia.

BACKGROUND OF THE INVENTION

Chest pain and ST segment alterations in the 12-lead electrocardiogram (ECG), have been the most important criteria for acute myocardial ischemia and evolving myocardial infarction, on which clinical decisions are based.

In the setting of acute myocardial ischemia knowledge of the severity and extent of the ischemia is of paramount importance. Whether the lack of blood supply in the acute situation is likely to induce permanent myocardial injury or whether it is more of a transient nature affects the choice and swiftness of treatment. In the acute situation the basis of the diagnosis lies on the symptoms and electrocardiography (ECG), which is generally the 12-lead ECG, possibly with some additional chest leads (V₄R, V₇, Vg). The ischemia-induced ECG alterations are described as either ST elevation or as non-specific ST changes, ST depression or T-wave inversion. Acute ST elevation, in combination with cardiac type chest pain, is an indication for rapid treatment, aiming for resolution of the suspected coronary occlusion, usually by either angioplasty or by thrombolysis. Nonspecific ST changes or an unaltered ECG leave more time for observation and medical therapy (Bertrand 2002). In acute coronary syndromes admission ECG changes have superior prognostic value to cardiac enzyme levels, regarding death or future myocardial infarction (MI) (Nyman 1993, Savonitto 1999). ECG is also immediately available, in contrast to cardiac enzyme levels, and allows for the triage of the patients. Unfortunately, in a small but undue amount of patients, who might benefit from prompt treatment, ischaemic changes in the 12-lead ECG go underestimated (Rouan 1989, Lee 1987, McCarthy 1993, Pope 2000, Lee 2000, Karlson 1991, Erhardt 2002). Mortality following a faulty discharge of acute MI (AMI) patients is doubled compared to those properly admitted (Lee 2000). Initial ECG is insufficient for the detection of acute coronary syndromes (Fesmire 1998, Karlson 1991, van De Werf 2003). Only 37% of the patients developing AMI show ST elevations on their admission ECGs (Karlson 1990). Yet, abnormal ECG remains the most important predictor of an acute coronary syndrome in a patient triage setting (Grijseels 1995). Several factors contribute to the outcome and recognition of significant ischemia: the location of the culprit occlusion and the resulting anatomical region of the ischemia, the severity of ischemia, duration of ischemia, bundle branch blocks, etc (van De Werf 2003).

Since the detection of acute coronary syndromes is heavily based on the initial ECG, any improvement in the diagnostic performance of ECG in acute myocardial ischemia is of paramount importance.

SUMMARY

It is an aim of this invention to provide a method and an arrangement for improved and simplified detection of myocardial ischemia, in particular acute myocardial ischemia. More specifically, it is an aim of this invention to improve the diagnostic performance of ECG.

It is also an aim of this invention to provide a method and an arrangement for distinguishing persons having injurious ischemia progressing to myocardial infarction and those having transient (non-injurious) ischemia.

It is also an aim of this invention to provide optimal recording locations for the ECG parameter.

These and other objects, together with the advantages thereof over known methods and arrangement are achieved by the present invention, as hereinafter described and claimed.

This invention is based on the discovery of ECG variables applicable for computer analysis and monitoring. According to this invention, the optimal variable for ischemia detection is QT integral. The QT integral can be applied to any number of ECG leads including the standard 12-lead ECG and is applicable for automatic computer analysis.

More specifically, the method is characterized by what is stated in claim 1 and the arrangement as stated in claim 10. According to this invention the QT integral already in single ECG lead is a sensitive and a relatively specific marker of acute myocardial ischemia. It is independent of ischemia location or underlying LVH. The QT integral exceeds in performance the conventional J amplitude criteria for myocardial ischemia. Furthermore, it is able to distinguish injurious from transient ischemia. Being detectable from a single lead, it is easily applicable into acute and monitoring situations. The optimal recording location is between standard V₄ and V₅ and therefore it may retrospectively be applied to 12-lead ECG databases. QT integral is easily applicable into acute and monitoring situations, and it is automatically analyzable.

The increase and elevation of the ST segment is a deviation which rapidly develops and disappears. It has been found that the increase of the ST level (point J level) disappears to 60 % already within a couple of hours. Therefore, the use of the QT integral is not as dependent on the actual time of the examination as the ST level and it can be used for information retrieval even if no increase in ST level has been found. Based on the above, the present invention provides for detection of myocardial ischemia and injury even when the ST segment elevation has resolved.

DESCRIPTION OF THE FIGURES

Figure 1 shows the BSPM layout.

Figures 2A and 2B show the optimal recording locations for acute ischemia detection. Small full dots mark the standard 12-lead ECG chest leads V₁-V₅. The full symbols mark the optimal recording locations for the J amplitude for each culprit artery occlusion: horizontal ellipse for left anterior descending coronary artery (LAD), vertical ellipse for right coronary artery (RCA), and rectangle for left circumflex coronary artery (LCX). The blank symbols mark the corresponding locations for the QT integral, with both RCA and LCX represented by a rectangle. The pentagon symbol marks the optimal recording location for the detection of ischemia by the QT integral in any culprit artery.

Figures 3 A and 3B show the group average QT integral values as a function of recording location. X-axis shows the lead label and Y-axis the QT integral value in mVms. The group average values for the patients approach zero, whereas the group average values for the controls deviate from the zero. Figures 4A and 4B show the group average spatial correlation of QRS and STT integrals in the patients and the controls, i.e., correlation of the QRS integral and STT integral with respect to the recording location on the torso.

DETAILED DESCRIPTION OF THE INVENTION

Definitions:

The QT integral represents the integral over the whole cardiac cycle, i.e. the mathematical sum of the QRS- and STT-integral values.

In body surface potential mapping (BSPM) 120 unipolar ECG leads, covering the whole thorax, and 3 limb leads are recorded simultaneously (Simelius 1996). Wilson's central terminal is used as a reference potential for all the chest leads. The signals are band-pass filtered and digitized.

The standard 12-lead ECG is comprised of 6 unipolar chest leads Vi-V₆, and 6 limb leads I, II, III and avR, avL, and avF, the latter mathematically derived from 3 recorded limb leads.

By transient ischemia is meant non- injurious ischemia (unstable angina pectoris, UAP).

By an injurious ischemia is meant an ischemia progressing to myocardial infarction.

By LVH patients are meant patients having aortic valve stenosis or arterial hypertension and the LVH may be diagnosed by echocardiography. The abbreviation "LVH" stands for left ventricular hypertrophy.

This invention provides a method for detecting in particular acute myocardial ischemia. The method comprises the steps: recording ECG-data from a person; transferring the ECG-data in digital form into a computer based automatic analysing system; calculating a QT integral from the ECG-data by using the computer based automatic analysing system; and receiving an interpretation of the QT integral values for detecting persons having acute myocardial ischemia.

This invention provides also an arrangement for detecting in particular acute myocardial ischemia in a person. The arrangement comprises: an input structure constructed to receive the person's ECG-data; and - a computer structure operatively connected to said input structure, adapted to calculate a QT integral from the ECG-data and interpreting the data by comparing it to the QT integral of control persons.

By a control person is meant a healthy person having no history of symptoms or signs suggestive of coronary artery disease. They show normal finding when examined with rest ECG and stress ECG or echocardiography.

A person to be studied may mean any person. However, if the QT integral is calculated from the ECG-data of a person having bundle branch block, atrial fibrillation, or pacemaker, the interpretation of the results is difficult. A person to be studied is usually a person having chest pain, the onset of chest pain being within less than 24 hours.

By a patient is meant in particular a person having ischemia, more specifically an injurious ischemia progressing to myocardial infarction or transient ischemia (unstable angina pectoris =UAP).

ECG- data is recorded from a person by using body surface potential mapping (BSPM) with any number of ECG leads.

The ECG-data is transferred in digital form into a computer based automatic analysing system. In addition a conventional electrocardiogram can be prepared from the data.

An input structure constructed to receive the person's ECG-data means the system by which the ECG data is collected from a person to be studied. It comprises the body surface potential mapping (BSPM) with any number of ECG leads recording the ECG-data. From this data is prepared a conventional electrocardiogram or this data is transferred in digital form to a computer based automatic analysing system or the data is used both for preparing an electrocardiogram and for transfer in digital form to a computer based automatic analysing system.

A computer structure operatively connected to said input structure means a computer structure adapted to automatically analyse the ECG data that it receives in digital form. The computer structure is adapted to calculate the QT integral from the ECG-data and interpret the data by comparing it to the QT integral of control persons. The QT integral is calculated over the whole cardiac cycle as the mathematical sum of the QRS and the STT integrals.

The arrangement comprises any number and location of ECG leads, including whole or part of ECG data recorded with body surface potential mapping (BSPM), and the standard 12-lead ECG and it's additional leads V₄ R, V₇ and V₈.

In a preferred embodiment of the invention the arrangement comprises alone or in any combination with other ECG leads a chest lead between standard V₄ and V₅.

The QT integral is usually measured in mVms values. The QT integral of the patients is approaching zero and the QT integral of healthy persons is deviating from zero. On the basis of patient group studies a person skilled in the art can set limit values or limit value ranges for QT integral for persons having injurious ischemia progressing to myocardial infarction, having transient ischemia (unstable angina pectoris =UAP) and healthy persons. In our studies on which this invention is based the QT integral values for control persons seem to be between values - 30 mVms and + 80 mVms. For patients the values were between values - 20 mVms and + 30 mVms (see Figure3). Depending on the number of persons studied the number values may be different. The variation range seems to be in the patient group equal or less than ¹A of the variation range in the control group. In severe cases the variation range may be in the patient group equal or less than 1/3 of the variation range in the control group.

The body surface potential mapping (BSPM) is preferably recorded from persons within 0- 24 hour, more preferably within 0- 12 hours from the onset of chest pain. The method and arrangement using QT integral is advantageously used to detect acute myocardial ischemia (within less than 24 hours from the onset of chest pain), although it can be used also to detect recent (4-30 days) or more remote ischemias (> 6 months). However, the significance of the method and arrangement of the invention is highest in acute cases.

As is disclosed here and exemplified in the examples the optimal single-lead variable for the detection of acute ischemia is the QT integral. The QT integral exceeds in performance both, the conventional 12-lead ECG based J-amplitude criteria for acute evolving myocardial infarction (van De Werf 2003, The joint eur 2000) and the J amplitude measured in its optimal chest location. Even the addition of a considerable number of LVH patients into the control group do not significantly deteriorate the performance of the QT integral, which may thus be considered a specific marker of ischemia. The T-apex amplitude performs equally in ischemia detection, but it may be considered non-specific, as it is likely to alter in a number of conditions. The QT integral decreases in ischemia, and even more in injurious ischemia. The optimal location for ischemia detection for the QT integral is between the standard leads V₄ and V₅. This optimal lead are in the vicinity of the optimal locations for the detection of ischemia induced by the occlusion of each of the three culprit arteries.

The detection of ischemia is generally not significantly different depending on the culprit artery, although the QT integral tends to detect LAD-occlusion induced ischemia better than RCA- or LCX-occlusion induced ischemia. It thus appears that our single lead parameter is independent of the location of ischemia.

The QT integral parameter is the sum of the QRS and STT integrals. As is disclosed here the patients have QT integral values approaching zero, as opposed to QT integral values in the controls diverging from zero. This phenomenon derives from the opposite polarity the QRS and the STT integrals, reflected by their negative spatial correlation in the ischemic patients contra parallel polarity in the controls. According to this disclosure a variable, the QT integral, detectable in a single lead, describes the phenomenon of divergence of the depolarization and repolarization wavefronts, and is able to detect acutely evolving myocardial injury more effectively than the conventional ECG markers. The difference in the polarity of the depolarization and repolarization integrals is akin to the QRS-T angle, which, in the setting of acute chest pain, is associated with increased long-term mortality risk, but interestingly, not with the final cardiac diagnosis (de Torbal 2004). The spatial QRS-T angle is able to identify patency of the infarct-related artery after thrombolysis, however with inferior ability relative to the conventional parameter of ST-segment resolution (Dilaveris 2005).

As is disclosed here later the largest deviation from zero for the J amplitude in the patients was observed in standard lead V₃. Yet, even in the pre -treatment patient group subsequently developing AMI, the largest group average J-amplitude deviation was only - 0.098 mV and did not distinguish these patients from the controls. In standard ECG this corresponds to only a 1 millimeter deviation, and would most likely go unnoticed in a clinical situation. Our automatic J amplitude analysis succeeded to detect acute ischemia in leads deviating from the standard 6 chest leads V₁-V₆. Nevertheless, even the automatically analyzed J amplitude in its optimal location failed in comparison with the QT integral. The absolute value of J amplitude performed similarly to the J amplitude, indicating that the direction of the deviation had no implication on the analysis.

The most sensitive and specific known ECG marker for acute myocardial infarction is the ST-segment elevation (Erhardt 2002), though the ST-segment elevation on admission ECG in suspected acute coronary syndrome has shown only 53% sensitivity and 95% specificity (Adams 1993). Our patients were, by inclusion criteria, exclusively ischemic and presented initially predominantly with acute ST elevation, as judged by the clinician, thus showing ECG signs of severe ischemia. Consequently, the clinically evaluated ST-elevation criteria in the initial ECG produced sensitivity of 72%. By the time of the BSPM recording, however, the sensitivity of the ST-elevation criteria had decreased to only 30%. Resolution of the ST-segment changes is considered a sign of withdrawing of ischemia (The joint Europ 2000). Yet, the QT integral was able to identify evolving myocardial infarction with a sensitivity of 90% and specificity of 50%, exceeding both the ST-elevation and ST- depression criteria. ST-segment elevation following angiographically successful revascularization indicates depressed coronary flow reserve (Feldman 2000), larger infarct size, and more severely impaired left ventricular (LV) function (Feldman 2000, van Hof 1997). Additionally, ST-segment elevation preceding and during angioplasty and early post angioplasty is related to the enzymatic infarct size and LV dysfunction despite comparable angiographic findings (Terkelsen 2006). Thus, electrocardiographic ST-segment elevation, pre- or post treatment has appeared, to date, the most sensitive indicator for evolving myocardial injury, challenged now by the single-lead QT-integral variable. In our patients the principal ST deviation, by the time of the BSPM recording, was ST depression. ST-T changes in an acute situation in combination with chest pain indicate worse prognosis (Nyman 1993, Scirica 2002). In the present study the QT integral was superior to the J amplitude, representing non-specific ST changes, in detecting injurious ischemia, in any chest location, let alone the standard 12-lead ECG criteria. Furthermore, the QT integral was, unlike the J amplitude, able to distinguish transient from injurious ischemia. ST depression of >0.2 mV in 3>leads in 12-lead ECG in acute coronary syndrome patients is associated with developing AMI (Lloyd- Jones 1998). In our patients the ST depression in the AMI and unstable angina pectoris (UAP) patients was only 0.1 mV. ST depression of 0.1 mV in (>1 mm) two or more contiguous leads is considered highly suggestive of an acute coronary syndrome (Bertrand 2002). This proved to be true in our patients, but falling behind the QT integral as a marker for acute coronary syndrome. ST depression in 12-lead ECG in patients with symptoms suggestive of AMI is a highly specific marker of MI and is related to poor prognosis (Lee 1993). The degree of the ST depression is more important marker of MI than the number of leads carrying the abnormality, indicating that the choice of the right parameter and selection of appropriate threshold value are more important than the abundance of information gained. Menown et al proved, in a study design very similar to ours, that a multivariate model constructed from three BSPM variables, exceeded the 12-lead ECG in the detection of acute myocardial infarction, producing sensitivity of 88% and specificity of 75% (Menown 2001). Our single-lead QT-integral parameter showed diagnostic performance of nearly the same level to their optimal multivariate model, which is constructed from three variables, each containing multivariate information from several BSPM leads. Our QT integral variable, in a single lead, was superior to their univariate ST elevation parameter, measured in any of the BSPM leads.

Our study group included no separate training and validation sets. To compensate for this, we calculated the sensitivity and specificity values with a cross validation method, mimicking the use of an independent validation set. Due to the inclusion criteria all of the patients presented with ST segment or T wave abnormalities in the initial ECG. The initial conventional ECG diagnosis thus produced a sensitivity of 100%. The patient selection was therefore not representative of acute chest-pain patients. In the BSPM analysis the failure of the conventional J amplitude criteria and J amplitude variable to properly identify acute ischemia may partly have been due to the timing of the BSPM recording. Though the patients were recruited within 12 hours from the onset of pain, the recording was performed in the hospital. Most of the patients had already received medical therapy (beta blockers, nitroglycerin, aspirin, morphin, oxygen etc.) before the BSPM recording, which most certainly had alleviated the ECG alterations. Yet, the same alleviation of the ECG changes also applies to the studied BSPM variables. The comparison of the variables was performed from a simultaneous recording to achieve a fair and commensurate assessment.

As is disclosed here the QT integral in a single ECG lead is a sensitive and specific marker of acute myocardial ischemia, challenging the conventional J amplitude. The QT integral in a single lead is superior to the J amplitude criteria in the detection of ischemia and ischemia progressing to MI, and is also able to distinguish between transient and injurious ischemia. The optimal recording location for the QT integral is between the standard leads V₄ and V₅. The ischemia location or addition of patients with left ventricular hypertrophy (LVH) into the control group had no substantial effect on the performance of the QT integral. QT integral is easily applicable into acute and monitoring situations, and it is automatically analyzable.

The optimal variables for ischemia detection is thus the QT integral. As is disclosed here later in more detail the QT integral in a single lead was superior to the J amplitude criteria (sensitivity of 85%, specificity of 36%) in the detection of ischemia (sensitivity of 89%, specificity of 38%) and ischemia progressing to myocardial infarction (sensitivity of 90%, specificity of 50%), and was also able to distinguish between transient and injurious ischemia (area under receiver operating characteristic curve (AUC) of 70%, p=0.034). The optimal recording location for the QT integral was between the standard leads V₄ and V₅. The ischemia location or addition of LVH patients into the control group had no substantial effect on the performance of the QT integral. Examples

We automatically analysed a 5 minute 120-lead BSPM recorded within 12 hours from the onset of chest pain in 79 patients with acute myocardial ischemia, in 84 healthy controls, and in 42 patients with LVH, with no coronary artery disease. Patients were grouped according to enzymatic myocardial damage (AMI, 68 patients) or not (UAP, 11 patients). Majority of the patients underwent coronary angiography indicated by the index event, and were grouped according to the culprit artery. Several variables and their optimal recording locations were automatically determined from the averaged BSPM recordings, in addition to the conventional J amplitude criteria for myocardial ischemia.

Patients were recruited at random dates during office hours at the Helsinki University Central Hospital. Inclusion criteria were chest pain and ST-segment abnormalities in the initial 12-lead ECG suggestive of myocardial ischemia or elevation of myocardial enzymes CK-Mb mass and/or TnT. Patients with bundle branch block, atrial fibrillation, or pacemaker were excluded. Total number of patients included in the final analysis was 79. The clinicians' initial 12-lead ECG interpretation was ST elevation (not necessarily fulfilling the criteria of amplitude and lead contiguity) in 57 patients and non-ST elevation (ST depression, T inversion, non-specific ST changes) in 22 patients. Thus, all of the patients presented with ECG abnormalities in the initial ECG. Study subject characteristics are presented in Table 1 and 2. Prior MI was present in 13 of the patients.

Patients were grouped according to whether they suffered myocardial damage (AMI) or not (UAP). CK-Mb mass>5 ug/1 was considered a sign of myocardial infarction. Additionally 4 patients with CK-Mb mass<5 ug/1 showed elevation in Troponin T (0.06- 0.13 ug/1). Nearly all of the patients (76) underwent coronary angiography indicated by the index event. All of the UAP patients showed significant (>50% of the luminal diameter) stenoses in the coronary angiography. The patients were grouped according to culprit artery stenosis, when the finding was unambiguous. Left anterior descending (LAD) was the culprit artery for 32 patients, right coronary artery (RCA) for 26 patients, and left circumflex coronary artery (LCX) for 10 patients. All of the patients had echocardiography performed acutely, within 12 hours from the onset of pain, to determine left ventricular ejection fraction (LVEF). The patients were further grouped according to whether the acute BSPM was recorded before (33 patients) or after (Post-treatment AMI, 45 patients) revascularisation. The patient group with BSPM before revascularisation was further divided into those who suffered myocardial damage (Pre -treatment AMI, 23 patients) and those with no myocardial damage (Pre-treatment UAP, 10 patients).

As controls were recruited 84 healthy volunteers with no history of symptoms or signs suggestive of coronary artery disease. They were examined with rest ECG and stress ECG or echocardiography, showing normal findings.

As an additional comparison group 42 patients with LVH were analysed (Oikarinen 2004). The LVH patients had aortic valve stenosis or arterial hypertension and the LVH was diagnosed by echocardiography. The patients showed LV mass indexed to body surface area of >116 g/m in men and >104 g/ m in women. They had no signs of coronary artery disease, no left ventricular dysfunction, and no pathological Q-waves in the 12-lead ECG.

All of the study subjects gave their written informed consent. The research protocol was approved by the local ethics committee and complied with the Declaration of Helsinki.

BSPM was recorded within 12(14) hours from the onset of chest pain. Resting BSPM for 5 minutes was recorded with 120 unipolar leads covering the whole thorax, and with 3 limb leads, as reported previously (Simelius 1996). Wilson's central terminal was used as a reference potential for all the chest leads. The electrodes were mounted on 18 strips, placed on the subject's thorax vertically with horizontal spacing determined by the dimensions of the upper body (Figure 1). After band-pass filtering from 0.16 Hz to 300 Hz, the signals were digitized with a sampling rate of 1000 Hz.

The BSPM data were visually inspected for validity of the recording, and signal-averaged according to criteria of 0.9 or greater correlation of the QRS complex to a selected template beat and maximum noise of 50 μV (Vaananen 2000). It was also required that the T wave fits an envelope of 110 - 200 μV around the template beat (Vaananen 2000). The baseline was defined from a 20 ms section of the TP segment. Invalid leads were replaced by interpolation of data from surrounding leads according to a method modified from the one presented by Oostendorp et al (Oostendorp 1989). The QRS onset and offset were determined automatically from the vector magnitude of a representative set of high-pass filtered leads. The automatic detection followed the basic guidelines given by Simson (Simson 1981). QRS was divided into 4 temporally equal segments, referred to as quartiles. T-wave end and apex were determined automatically, as described earlier (Oikarinen 1998). The median of T-wave end of all leads was used for calculations. STT segment was defined as the time interval from the QRS offset to the T-wave offset. Time integrals were calculated for the whole QRS, the QRS quartiles, and the STT segment. QT integral represented the integral over the whole systolic cycle, i.e. the sum of QRS and STT integrals.

Discriminant indexes (DI) were calculated as described by Kornreich et al (Kornreich 1991). DI identified the optimal recording locations to separate MI patients from the controls. The mean value of the control group was subtracted from the corresponding mean of the MI patient group. The difference was divided by the pooled standard deviation (SD) of the both groups.

The DI value is directly proportional to the t- value in the student's t-test. The greater the absolute DI value ([DI]), the better the ability of the lead to differentiate between the two groups. Negative DI values indicate lower and positive DI values higher parameter values for the patient group as compared to the control group.

Continuous values are presented as mean ± SD. Statistical significance of difference between the groups was determined with the Mann- Whitney U test. Correlations between the parameters were examined with Pearson's correlation coefficient (r). Receiver operating characteristic (ROC) curves were created to assess the performance of the parameters in the optimal leads. Areas under ROC curves (AUC) derived from the same sets of cases were compared by a method described by Hanley (Hanley 1983). AUCs of ROC curves derived from different sets of cases were compared by a simpler formula, leaving out the correlative nature of the data. A 2 tailed p-value ≤ 0.05 was considered statistically significant. The Statistical Package for the Social Sciences (SPSS, Inc., Chicago, IL, USA) for Windows (version 10.0) biostatistics software was used. Sensitivities and specificities were also calculated with leave-one-out cross-validation (Fugunaka 1972). Briefly, each case was classified in turn according to the 95 % specificity threshold calculated from all the other cases. Thus, each test case is independent of the learning set.

The optimal variables for detecting acute ischemia

The optimal variables for detecting acute myocardial ischemia, as compared to the healthy controls, were the T-apex amplitude with AUC in the optimal lead of 85% (p<0.001), and the QT integral with AUC of 85% (p<0.001), followed by the STT integral with AUC of 83% (p<0.001) (Table 3). The optimal lead for the QT integral in ischemia detection was located between the standard leads V₄ and V₅ (Figure 2). The AUC for the QT integral was larger (p=0.0078) than for the J amplitude in its optimal lead on the lower abdomen.

When LVH patients were included in the control group, the AUCs deteriorated in the previously defined optimal leads, but the difference between the two AUCs was not significant for the QT integral (p=0.484). When the optimal lead was defined anew, in comparison with the control group including LVH patients, the AUC was 82% (p<0.001, lead 67) for the QT integral.

Of the QRS-integral quartiles the 1^st and the 2^nd performed best in detecting ischemia (Table 2). Adding the LVH group into the control group did not significantly affect the performance of the first two QRS integral quartiles.

Table 1

Baseline characteristics of the study subjects

BSPM Number of Sex F/M age BMI LVEF CK Mb max subjects

Controls 84 18/66 53±13 25.7±3.7 >50

LVH patients 42 17/25 63±12 26.3±3.6 >50 Acute 79 22/57 61±11 27.5±54.4 53±10 147±193

AMI 68 18/50 61±11 27.6±4.4 52±10 171±198

UAP 11 4/7 62±10 26.5±4.3 60±6 3±1 BMI = body mass index, LVEF = left ventricular ejection fraction Table 2

Characteristics of ischemia and conventional ECG markers. The initial ECG is the 12-lead ECG acquired in the first clinical contact with the patient, judged by the clinician. J- amplitude criteria include both elevation and depression and are defined from a 12-lead ECG derived from the BSPM registration.

Patients Controls

Time (h) pam-BSPM 7.3±3.0

Time (h) pain-treatment 6.2±8.8

Initial ECG 57

ST elevation

Initial ECG non ST 22 elevation

J amplitude criteria 67 54 positive

J amplitude criteria 12 30 negative

J amplitude elevation 24 18

Table 3

Variables and their optimal locations for detecting myocardial ischemia

AUC (CI) % AUC (CI) % MI vs DI Optimal

MI vs Controls Controls and LVH lead

T apex amplitude 84.7 (78.8-90.6) 77.7 (71.3-84.2) -1.170 71

QT integral 84.6 (78.6-90.6) 81.6 (75.6-87.6) -1.124 68

STT integral 83.4 (77.1-89.6) 74.1 (67.2-81.0) -1.087 71

QRS integral 74.6 (67.2-82.1) 65.6 (58.2-73.0) -0.827 42

1^st QRS integral 79.4 (72.5-86.3) 75.5 (68.5-82.5) 1.019 104

2^nd QRS integral 78.5 (71.5-85.4) 79.7 (73.4-85.8) 0.984 110

3^rd QRS integral 68.9 (60.7-77.1) 55.0 (47.6-63.5) -0.639 34

4^th QRS integral 66.4 (58.2-74.6) 60.9 (53.2-68.5) -0.567 55

J amplitude 71.1 (63.2-79.1) 69.9 (62.6-77.2) -0.667 34

[J amplitude] 72.3 (64.4-80.3) 67.1 (59.4-74.8) 0.718 33

AUC = area under receiver operating characteristic curve, CI = confidence interval The conventional amplitude variable

The conventional J amplitude criteria for myocardial ischemia, including both ST elevation and non-ST elevation criteria, (The Joint Europ 2000) derived from the BSPM leads, including limb leads, corresponding to the 12-lead ECG, produced a sensitivity of 85% and a specificity of 36% (Tables 2 and 4). The J amplitude elevation criteria only produced a sensitivity of 30% and a specificity of 79%. The sensitivity and specificity values for the detection of ischemia for the QT integral are shown in Table 4, for the different patient groups. The sensitivity and specificity values for the QT integral in the AMI and post- treatment AMI patient groups were superior to those produced by the conventional J- amplitude criteria. The pre-treatment AMI group also showed superior specificity, with equal sensitivity, to the conventional J-amplitude criteria. The only group falling behind the J-amplitude criteria in sensitivity and specificity values was the UAP patient group. The J amplitude in its optimal lead was inferior to the QT integral in detecting ischemia and infarction (Tables 3 and 5). The absolute value of J amplitude, [J amplitude], performed similarly to the J amplitude.

Table 4

Sensitivities and specificities of the conventional J amplitude criteria in 12-lead ECG and of the single-lead QT integral for ischemia detection. The sensitivities and specificities of the QT integral are calculated by a cross validation method.

Sensitivity % Specificity %

J amplitude criteria, all ischemia 85 36 patients

J amplitude criteria, AMI patients 85 36

ST elevation criteria, all ischemia 30 79 patients

QT integral

All ischemia patients 89 38

AMI patients 90 50

Pre-treatment AMI patients 83 50

Post-treatment AMI patients 87 51

Pre-treatment UAP patients 80 32 The group average J amplitude showed, of the BSPM leads, the largest deviation from zero in the control group in lead 40, below standard V₂ (-0.049±0.094 mV), and in the patient group in the lead 47, closely corresponding to standard V₃ (-0.111±0.135 mV). The patient group differed from the controls (p<0.001). On the anterior torso the average J amplitude values and corresponding DI values were negative, indicating absolute ST depression and ST depression greater for the patient group relative to the control group. In the AMI patient group the largest group average J-amplitude deviation was found in lead 47 (-0.116±0.141 mV) and in the UAP patient group in lead 40 (-0.082±0.083 mV). The AMI patient group differed from the controls (p<0.001), but the UAP patient group showed no difference from the controls (p=0.159).

In the pre-treatment AMI group the largest group average J-amplitude deviation was found in lead 47 (-0.098±0.154 mV), and in pre-treatment UAP and post-treatment AMI groups in lead 40 (-0.094±0.082 mV and -0.127=1=0.151 mV, respectively). The post-treatment AMI group differed from the control group (p<0.001), but the pre-treatment AMI or the pre-treatment UAP groups showed no difference as compared to the controls (p=0.062 and p=0.082, respectively).

AMI vs UAP

The QT integral, the STT integral, and the QRS integral tended to detect AMI better than UAP, but differences between the two AUCs were non-significant (Table 5). The QT integral and QRS integral were, however, able to distinguish between the two patient groups in leads deviating from the optimal ones (AUCs of 70%, p=0.034, and 77%, p=0.004, respectively).

Table 5

Variables and their optimal locations for detecting acute myocardial infarction (AMI) and unstable angina pectoris (UAP)

TUC (CI) % MI vs AUC (CI) % MI vs Optimal

Controls Controls and LVH lead

QT integral AMI ^~87T(8L5^92!9) 84^(783^900) 68^~ QT integral UAP 81.9 (72.6-91.2) 77.7 (69.1-86.3) 34

STT integral AMI 83.8 (77.4-90.2) 74.4 (67.6-81.7) 71

STT integral UAP 78.1 (66.3-89.9) 78.4 (67.1-89.6) 60

QRS integral AMI 75.6 (68.0-83.2) 66.7 (59.1-74.3) 42

QRS integral UAP 69.8 (52.2-87.4) 54.5 (38.3-70.8) 13

J amplitude AMI 71.4 (63.0-79.8) 70.1 (62.4-77.8) 34

J amplitude UAP 69.0 (53.0-85.1) 68.9 (54.2-83.6) 28

[J amplitude AMI] 74.3 (66.2-82.4) 69.1 (61.2-77.0) 33

[J amplitude UAP] 61.1 (40.9-81.4) 59.2 (39.2-79.2) 57

AUC = area under receiver operating characteristic curve, CI = confidence interval

The AUC, in the optimal lead, of the QT integral was larger than that of the J amplitude for the AMI group (p<0.002) but not for the UAP group (p=0.1052). J amplitude detected AMI and UAP equally (AUCs of 71% and 69%, respectively, p=0.7949 for the difference between the AUCs), and was unable to distinguish between the two (AUC 56%, p=0.591).

Performance of the variables with respect to treatment in the acute phase The QT integral was the best variable in detecting ischemia progressing into MI, both in the pre-treatment and post-treatment AMI groups (Table 6). The QT integral, and the STT integral tended to identify injurious ischemia better than UAP (p=NS). The QT integral produced larger AUCs than the J amplitude in the pre-treatment AMI and post-treatment AMI (p<0.0020, p=0.0536, respectively) patient groups.

Table 6

AUCs for detecting acute ischaemia by the optimal BSPM leads and the optimal standard

12-lead ECG lead (V₃) for the J amplitude in pre- and post-treatment AMI and UAP patients

Lead Pre-treatment Post-treatment Pre-treatment UAP

AMI AUC (CI) % AMI AUC (CI) % AUC (CI) % _____ ——--———— ^ _-_-__--_-_-_-___^^

QT integral 68 85.8 (76.6-95.1) 87.9 (81.5-94.3) 72.9 (57.9-87.8) STT integral 71 82.1 (71.6-92.7) 84.7 (77.6-91.7) 79.0 (64.8-93.3)

Culprit artery in acute ischemia

The three optimal parameters were compared with the conventional J amplitude in detecting ischemia induced by specific culprit artery occlusion. For the J amplitude the optimal leads for the different culprit arteries formed a line running perpendicular to the cardiac anatomical axis (Figure 2).

The QT integral detected acute ischemia induced by occlusion of any culprit artery with larger AUCs than the J amplitude (Table 7). The optimal leads for the QT integral for detecting LAD occlusion induced ischemia lay on the 4^th intercostal space above V₄ and V₅, and for RCA and LCX induced ischemia in the place OfV₆ (Figure 2). The AUC in LAD occlusion for the QT integral (91%) was significantly larger than the AUC for the J amplitude (69%) (p<0.002) in the optimal lead, but the difference between the AUCs in RCA occlusion and LCX occlusion induced ischemia for the QT integral and for the J amplitude in the optimal leads failed to reach significance (p=0.6101 and p=0.4295, respectively). In the optimal lead (68) for ischemia detection for all the patient groups the QT integral produced AUCs of 90%, 80%, and 81% for LAD, RCA, and LCX occlusion, respectively. The differences between the AUCs for different culprit arteries were non- significant in lead 68.

Table 7

AUCs for detecting acute ischaemia induced by LAD, RCA, and LCX as culprit arteries in optimal leads for the whole patient group (superior rows) and for each culprit artery (inferior rows).

Lead^''"''AUc^"(CΪy% Lea d ^"'" AUC (C ϊ) Lead AUC (CI) %

% J amplitude 35 68.5 65 77.2 48 73.8 (57.8-79.1) (66.9-87.6) (57.9-89.7)

J amplitude 34 69.9 34 74.2 34 70.6

(59.0-80. 8) (62.7-85 .8) (52.5-88.7)

QT integral 67 90.8 76 80.6 76 87.9

(85.1-96.5) (70.9-90.3) (76.1-99.6)

QT integral 68 90.3 68 79.6 68 80.6

(84.1-96.6) (68.9-90.3) (65.8-95.3)

The group average QT integral absolute values were almost consistently (in 97/120 leads) decreased in patients relative to the controls, in other words, the QT integral values in patients approached zero, whereas in the controls the QT integral values departed from zero (Figure 3). Values larger in patients than in controls covered a belt-like area running from right inferior flank to the left shoulder, perpendicular to the cardiac anatomical axis. QT integral values were smaller in the patients than in the controls on the left side of the torso with a maximum difference between the groups on the left flank area. QT integral values in AMI patients were smaller than in the UAP patients on the left side of the torso, with a maximum difference on the left flank area.

The group average QRS and STT integrals showed negative correlation with respect to recording location, i.e. negative spatial correlation, in the patient group (r=-0.435, p<0.001), and positive correlation in the control group (r=0.300, p=0.001) (Figure 4). The QT integral showed weak correlations with CK-Mb mass in 15 chest leads and with left ventricular ejection fraction in 25 chest leads, with correlations in the optimal lead 68 of r=-0.224 (p=0.047), and r=0.233 (p=0.046), respectively. No correlation was observed between the CK-Mb mass max and left ventricular ejection fraction.

Acute cardiac chest pain: All patients

This group has both patients with UAP without permanent myocardial damage and patients with AMI and permanent damage. For patients with the onset of chest pain less than 12 hours before, the following parameters detected acute, potentially injurious myocardial ischemia in the following order of superiority: QT integral (the best parameter), STT integral, QT area, QRS integral and the conventional ST60 amplitude (Table 6).

For patients with the onset of chest pain 5 to 30 days before the parameters performed in the following order of superiority: STT integral (the best), QT integral, QT area, QRS integral and the conventional ST60 amplitude (Table 8).

Table 8.

Variables and their optimal locations for detecting acute ischemia (age < 30 days)

Time from the onset of ischemia AUC (CI) %

Ischemia patients vs Controls QT integral < 12 hours 84.6 (78.6-90.6)

5-30 days 77.3 (69.5-85.2)

STT integral < 12 hours 83.4 (77.1-89.6)

5-30 days 77.8 (70.1-85.4)

QT area < 12 hours 80.0 (73.3-86.6)

5-30 days 74.4 (66.5-82.3)

QRS integral < 12 hours 74.6 (67.2-82.1)

5-30 days 71.9 (64.0-79.8)

ST60 amplitude < 12 hours 68.6 (60.5-76.7)

5-30 days 67.6 (59.2-75.9)

AUC = area under the receiver operating characteristic curve, CI = confidence interval

Performance of the tested variables in subgroups of patients according to the presence of acute myocardial infarction (AMI) or unstable angina (UAP)

The parameters detected AMI both in patients with the onset of chest pain < 12 hours earlier and 5-30 days earlier with following order of superiority : QT integral (the best parameter), STT integral, QT area, QRS integral and the conventional ST60 amplitude

(Table 9).

The parameters detected acute UAP with following order of superiority: QT integral (the best parameter), STT integral, QT area, conventional ST60 amplitude and the QRS integral (Table 2). For patients with UAP with the onset of chest pain 5-30 days earlier, the conventional ST60 amplitude and STT integral outperformed the QT integral, QT area and the QRS integral (Table 9).

Table 9.

Variables and their optimal locations for detecting acute myocardial infarction (AMI) and unstable angina (UAP) (age < 30 days)

< 12 hours 5-30 days

AUC (CI) % Optimal lead AUC (CI) % Optimal lead

QT integral AMI 87.2 (81.5-92.9) 68 79.6 (71.5-87.6) 68 QT integral UAP 81.9 (72.6-91.2) 34 71.7 (58.3-85.1) 34 STT integral AMI 83.8 (77.4-90.2) 71 78.3 (70.2-86.4) 68

STT integral UAP 78.1 (66.3-89.9) 60 72.9 (56.3-89.4) 67

QT area AMI 81.0 (74.3-87.7) 111 77.8 (70.0-85.5) 7

QT area UAP 77.8 (65.7-90.0) 15 69.0 (52.9-85.2) 36

QRS integral AMI 75.6 (68.0-83.2) 42 73.6 (65.5-81.6) 49

QRS integral UAP 69.8 (52.2-87.4) 13 52.2 (31.8-72.7) 72

ST60 AMI 68.8 (60.3-77.3) 81 67.4 (58..7-76.1) 66 ST60 UAP 77.9 (63.8-92.1) 71 74.9 (60.0-89.7) 60

Recent and remote myocardial infarction (MI)

In the group of patients with recent (4-30 days) and remote (> 6 months) MI the tested parameters performed in the similar fashion to the acute injury patients. In the order of superiority the best variable was QT integral, followed by STT integral, QT area, and QRS integral (Table 10).

Table 10. Variables and their optimal locations for detecting recent myocardial infarction (MI)

MI age MI age

4-30 days > 6 months

AUC (CrJ% όplunaϊTead AUC^"(CI) % Optirnalleal^"

QT integral 90.2 (85.2-95.1) 69 88.2 (83.0-93.4) 62

STT integral 87.8 (82.1-93.6) 72 89.7 (84.9-94.4) 76

QT area 86.3 (80.2-92.5) 111 83.3 (77.0-89.6) 110

QRS integral 86.1 (79.6-92.6) 27 83.3 (77.1-89.5) 34

References

1. Bertrand ME, Simoons ML, Fox KAA, Wallentin LC, Hamm CW, McFadden E, De Feyter PJ, Speccihia G, Ruzyllo W. Management of acute coronary syndromes in patients presenting without persistent ST-segment elevation. The Task Force on the Management of Acute Coronary Syndromes of the European Society of Cardiology .

Eur Heart J 2002;23: 1809-1840.

2. Rouan GW, Lee, TH, Cook EF, Brand DA, Weisberg MC, Goldman L. Clinical characteristics and outcome of acute myocardial infarction in patients with initially normal or nonspecific electrocardiograms (a report from the Multicenter Chest Pain Study). Am J Cardiol 1989; 64:1087-92.

3. Nyman I, Areskog M, Areskog NH, Swahn E, Wallentin L. Very early risk stratification by electrocardiogram at rest in men with suspected unstable coronary heart disease. The RISC Study Group. J Intern Med 1993; 234:293-301.

4. Savonitto S, Ardissino D, Granger CB, Morando G, Prando MD, Mafrici A, Cavallini C, Melandri G, Thompson TD, Vahanian A, Ohman EM, Califf RM, Van de Werf F,

Topol EJ. Prognostic value of the admission electrocardiogram in acute coronary syndromes. JAMA 1999; 281;707-13.

5. Fesmire FM, Percy RF, Bardoner JB, Wharton DR, Calhoun FB. Usefulness of automated serial 12-lead ECG monitoring during the initial emergency department evaluation of patients with chest pain. Ann Emerg Med 1998; 31:3-11.

6. Lee TH. Rouan GW, Weisberg MC, Brand DA, Cook EF, Acampora D, Goldman L. Sensitivity of routine clinical criteria for diagnosing myocardial infarction within 24 hours of hospitalization. Ann Intern Med 1987; 106: 181-6.

7. McCarthy BD, Beshansky JR, D'Agostino RB, Selker HP. Missed diagnoses of acute myocardial infarction in the emergency department: results from a multicenter study.

Ann Emerg Med 1993;22:579-82.

8. Pope JH, Aufderheide TP, Ruthazer R, Woolard RH, Feldman A, Beshansky JR, Griffith JL, Selker HP. Missed diagnoses of acute cardiac ischemia in the emergency department. N Engl J Med 2000; 342:1163-70. 9. Lee TH, Goldman L. Evaluation of the patient with acute chest pain. N Engl J Med

2000;342:l 187-95.

10. Karlson BW, Herlitz J, Edvardsson N, Emanuelsson H, Sjolin M, Hjalmarson A. Eligibility for intravenous thrombolysis in suspected acute myocardial infarction. Circulation 1990; 82:1140-6. 11. Karlson BW, Herlitz J, Wiklund O, Richter A, Hjalmarson A. Early prediction of acute myocardial infarction from clinical history, examination and electrocardiogram in the emergency room. Am J Cardiol 1991 ;68: 171-5.

12. Grijseels EW, Deckers JW, Hoes AW, Hartman JA, Van der Does E, Van Loenen E, Simoons ML. Pre -hospital triage of patients with suspected myocardial infarction.

Evaluationn of previously developed algorithms and new proposals. Eur Heart J 1995; 16:325-32.

13. Oikarinen L, Karvonen M, Viitasalo M, Takala P, Kaartinen M, Rossinen J, Tierala I, Hanninen H, Katila T, Nieminen M, Toivonen L. Electrocardiographic assessment of left ventricular hypertrophy with time-voltage QRS and QRST-wave areas. Journal of

Human Hypertension 2004; 18, 33-40.

14. Simelius K, Tierala I, Jokiniemi T, Nenonen J, Toivonen L, Katila T. A body surface mapping system in clinical use. Med Biol Eng 1996;34 (suppl):107-108.

15. Vaananen H, Korhonen P, Montonen J, Makijarvi M, Nenonen J, Oikarinen L, Simelius K, Toivonen L, Katila T. Non- invasive arrhythmia risk evaluation in clinical environment. Herzschr Elektrophys 2000; 11:229-234.

16. T.F. Oostendorp, A. van Oosterom, and G. Huiskamp. Interpolation on a triangulated 3d surface. J. Comp. Physicsl989;80:331-343.

17. M.B. Simson. Use of signals in the terminal qrs complex to identify patients with ventricular tachycardia after myocardial infarction. Circulation 1981;64:235-42.

18. Oikarinen L, Paavola M, Montonen J, Viitasalo M. Makijarvi M. Toivonen L. Katila T. Magnetocardiographic QT interval dispersion in postmyocardial infarction patients with sustained ventricular tachycardia: Validation of automated QT measurements. Pacing Clin Electrophysiol 1998;21: 1934-1942. 19. Kornreich F, Montague TJ, Rautaharju P. Identification of first acute Q wave and non- Q wave myocardial infarction by multivariate analysis of body surface potential maps. Circulation 1991:84:2442-2453.

20. Hanley J, McNeil B. A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology 1983;148:839-43. 21. Wong GC, Frisch D, Murphy SA, Sabatine MS, Pai R, James D, Kraimer N,

Katsiyannis PT, Marble SJ, DiBattiste PM, Demopoulos LA, Gourlay SG, Barron HV, Cannon CP, Gibson CM. Time for contrast material to traverse the epicardial artery and the myocardium in ST-segment elevation acute myocardial infarction versus unstable angina pectoris/non-ST-elevation acute myocardial infarction. Am J Cardiol

2003; 91:1163-7.

22. Lloyd- Jones DM, Camargo CA Jr, Lapuerta P, Giuliano RP, O'Donnell CJ. Electrocardiographic and clinical predictors of acute myocardial infarction in patients with unstable angina pectoris. Am J Cardiol 1998; 81:1182-6.

23. Scirica BM, Cannon CP, McCabe CH, Murphy SA, Anderson HV, Rogers WJ, Stone PH, Braunwald E. Thrombolysis in Myocardial Ischemia III Registry Investigators. Am J Cardiol 2002; 90:821-6.

24. The Joint European Society of Cardiology/ American College of Cardiology Committee. Myocardial infarction redefined - A consensus document of The Joint

European Society of Cardiology/ American College of Cardiology Committee for the Redefinition of Myocardial Infarction. Eur Heart J 2000;21: 1502-1513.

25. Van de Werf F, Ardissino D, Betriu A, Cokkinos DV, FaIk E, Fox KAA, Julian D, Lengyel M, Neumann F-J, Ruzyllo W, Thygesen C, Underwood SR, Vahanian A, Verheugt FWA, Wijns W. Management of acute myocardial infarction in patients presenting with ST-segment elevation. The Task Force on the Management of Acute Myocardial Infarction of the European Society of Cardiology. Eur Heart J 2003; 24:28- 66.

26. Erhardt L, Herlitz J, Bossaert L, Halinen M, Keltai M, Koster R, Marcassa C, Quinn T, van Weert H. Task force on the management of chest pain. Eur Heart J 2002; 23 : 1153-

1176.

27. Adams J, Trent R, Rawles J. Earliest electrocardiographic evidence of myocardial infarction: Implications for thrombolytic treatment. BMJ 1993;307:409-413.

28. Terkelsen CJ, Norgaard BJ, Lassen JF, Poulsen SH, Gerdes JC, Sloth E, Gotzshe LBH, Romer FK, Thuesen L, Nielsen TT, Andersen HR. Potential significance of spontaneous and interventional ST-changes in patients transferred for primary percutaneous coronary intervention: observations from the ST-MONitoring in Acute Myocardial Infarction study (The MONAMI study). Eur Heart J 2006; 27:267-275.

29. Feldman LJ, Himbert D, Juliard JM, Karrillon GJ, Benamer H, Aubry P, Boudvillain O, Seknadji P, Faraggi M, Steg G. Reperfusion syndrome: Relationship of coronary blood flow reserve to left ventricular function and infarct size. J Am Coll Cardiol 2000; 35:1162-1169. 30. van Hof AWJ, Liem A, de Boer M-J, Zijlstra F, for the Zwolle Myocardial Infarction

Study Group. Clinical value of 12-lead electrocardiogram after successful reperfusion therapy for acute myocardial infarction. Lancet 1997; 350:615-619.

31. De Torbal A, Kors JA, van Herpen G, Meij S, Nelwan S, Simoons ML, Boersma E. The electrical T-axis and the spatial QRS-T angle are independent predictors of long-term mortality in patients admitted with acute ischemic chest pain. Cardiology 2004; 101:199-207.

32. Dilaveris P, Anastasopoulos A, Androulakis A, Theoharis A, Zumerle B, Tzannetis G, Kallikazaros I, Stefanadis C. Effects of thrombolysis on vectorcardiographic indices of ventricular repolarization: correlation with ST-segment resolution. J Electrocardiol

2005; 38:347-53.

33. Lee HS, Cross SJ, Rawles JM, Jennings KP. Patients with suspected myocardial infarction who present with ST depression. Lancet 1993; 342:1204-07.

34. Menown IBA, Allen J, McC Anderson J, Adgey AAJ. ST depression only on the initial 12-lead ECG: early diagnosis of acute myocardial infarction. Eur Heart J 2001;

22:218-227.

35. Keinosuke Fugunaka. Introduction to Statistical Pattern Recognition. Academic Press, New York 1972:150.

Claims

1. Method for detecting myocardial ischemia in a person, which comprises the steps: recording ECG- data from the person; - transferring the ECG-data in digital form into a computer based automatic analysing system; calculating a QT integral from the ECG-data by using the computer based automatic analysing system; and receiving an interpretation of the QT integral values for detecting persons having myocardial ischemia.

2. The method according to the claim 1, wherein the method is used for detecting acute myocardial ischemia.

3. The method according to the claim 1 or 2, wherein ECG-data is recorded whole or part with body surface potential mapping (BSPM).

4. The method according to the claim 1 or 3, wherein the ECG-data is recorded from any location of ECG lead(s).

5. The method according to any one of claims 1 to 4, wherein the ECG-data is recorded from the standard 12-lead ECG and optionally it's additional leads V₄R,

6. The method according to any one of the preceding claims, wherein the ECG comprises a chest lead between standard V₄ and V₅.

7. The method according to any one of the preceding claims, wherein the QT integral of the persons having ischemia is approaching zero and the QT integral of control persons is deviating from zero.

8. The method according to any one of the preceding claims, wherein the body surface potential mapping (BSPM) is recorded from patients within 0- 24 hour, preferably within 0- 12 hours from the onset of chest pain.

9. The method according to any one of the preceding claims, wherein the QT integral is calculated over the whole cardiac cycle as the mathematical sum of the QRS and the STT integrals.

10. The method according to any one of the preceding claims, wherein the computer based automatic analysing system for ECG-data calculates the integral from the ECG-data and interprets the values by distinguishing persons having injurious ischemia progressing to myocardial infarction, having transient ischemia (unstable angina pectoris =UAP) and healthy persons.

11. An arrangement for detecting myocardial ischemia in a person, which comprises: an input structure constructed to receive the person's ECG-data; and a computer structure operatively connected to said input structure, adapted to calculate a QT integral from the ECG-data and interpreting the data by comparing it to the QT integral of control persons.

12. The method according to the claim 11, wherein the arrangement is used for detecting acute myocardial ischemia.

13. The arrangement according to claim 12, wherein ECG-data is recorded whole or in part with body surface potential mapping (BSPM).

14. The arrangement according to claim 11 or 13, wherein the ECG-data is recorded from any location of ECG leads.

15. The arrangement according to any one of claims 11 to 14, wherein the ECG-data is recorded from the standard 12-lead ECG and optionally it's additional leads V₄R, V₇ and V₈.

16. The arrangement according to any one of claims 11 to 15, wherein the ECG-data is recorded from a chest lead between standard V₄ and V₅.

17. The arrangement according to any one of claims 11 to 16, wherein the QT integral of the persons having ischemia is approaching zero and the QT integral of control persons is deviating from zero.

18. The arrangement according to any one claims 11 to 17, wherein the body surface potential mapping (BSPM) is recorded from patients within 0- 24 hour, preferably within 0- 12 hours from the onset of chest pain.

19. The arrangement according to any one of claims 10 to 18, wherein the QT integral is calculated over the whole cardiac cycle as the mathematical sum of the QRS and the STT integrals.

20. The arrangement according to any one of claims 11 to 19, wherein the computer structure having automatic analysing system for ECG-data calculates the integral from the ECG-data and interprets the values by distinguishing persons having injurious ischemia progressing to myocardial infarction, having transient ischemia (unstable angina pectoris =UAP) and healthy persons.