* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Text - Ulster Institutional Repository
Survey
Document related concepts
Transcript
Epicardial Potentials computed from the Body Surface Potential Map using inverse electrocardiography and an individualised torso model improve sensitivity for Acute Myocardial Infarction diagnosis. Michael J Daly MB MRCP a, Dewar D Finlay PhD b, Daniel Guldenring PhD b, Raymond R Bond PhD b, Aaron J McCann PhD c, Peter J Scott MD MRCP a, Jennifer A Adgey MD DSc FRCP a and Mark T Harbinson MD FRCP c. a. The Heart Centre, Royal Victoria Hospital, Grosvenor Road, Belfast, Northern Ireland UK b. School of Computing and Mathematics and Computer Science Research Institute, University of Ulster, Northern Ireland, UK c. Centre for Vision and Vascular Sciences, Queen’s University, Whitla Medical Building, 97 Lisburn Road, Belfast, Northern Ireland UK Word Count: 3,616 Keywords: inverse electrocardiography; body surface potential mapping; acute myocardial infarction; body mass index Financial support: Dr Michael J Daly is supported by The Heart Trust Fund (Royal Victoria Hospital), 9B Castle Street, Comber, Newtownards Northern Ireland BT23 5DY Relationship with Industry/Conflict of Interest: None Address for correspondence: Professor Jennifer A. Adgey The Heart Centre, Royal Victoria Hospital, Grosvenor Road, Belfast Northern Ireland BT12 6BA Email: [email protected] Tel: +44 2890 633714 Fax: +44 2890 635212 1 Abstract Introduction: Epicardial potentials (EP) derived from the body surface potential map (BSPM) improve acute myocardial infarction (AMI) diagnosis. In this study, we compared EP derived from the 80-lead BSPM using a standard thoracic volume conductor model (TVCM) with those derived using a patient-specific torso model (PSTM) based on body mass index (BMI). Methods: Consecutive patients presenting to both the ED and pre-hospital coronary care unit between August 2009 and August 2011 with acute ischaemic-type chest pain at rest were enrolled. At first medical contact, 12-lead ECG and BSPM were recorded. BMI for each patient was calculated. Cardiac troponin-T (cTnT) was sampled 12h after symptom onset. Patients were excluded from analysis if they had any electrocardiographic confounders to interpretation of the ST-segment. A cardiologist assessed the 12-lead ECG for STEMI by Minnesota criteria and the BSPM. BSPM ST-elevation (STE) was ≥0.2mV in anterior, ≥0.1mV in lateral, inferior, RV or high right anterior and ≥0.05mV in posterior territories. To derive EP, the BSPM data were interpolated to yield values at 352-nodes of a Dalhousie torso. Using an inverse solution based on the boundary element method, EP at 98 cardiac nodes positioned within a standard TVCM were derived. The TVCM was then scaled to produce a PSTM, using a model developed from computed tomography (CT) in 48 patients of varying BMI, and EP re-calculated. EP >0.3mV defined STE. A cardiologist blinded to both the 12-lead ECG and BSPM interpreted the EP map. AMI was defined as cTnT ≥0.1µg/L. Results: Enrolled were 400 patients (age 62 ± 13 yrs; 57% male): 80 patients had exclusion criteria. Of the remaining 320 patients, BMI was 27.8 5.6kg m-2. Of these, 180 (56%) had 2 AMI. Overall, 132 had Minnesota STE on ECG (sensitivity 65%, specificity 89%) and 160 had BSPM STE (sensitivity 81%, specificity 90%). EP STE occurred in 165 patients using TVCM (sensitivity 88%, specificity 95%, p<0.001) and in 206 patients using PSTM (sensitivity 98%, specificity 79%, p<0.001). Of those with AMI by cTnT and EP ≤0.3mV using TVCM (n=22), 18 (82%) patients had EP >0.3mV when an individualised PSTM was used. Conclusion: Among patients presenting with ischaemic-type chest pain at rest, EP derived from BSPM using a novel PSTM significantly improves sensitivity for AMI diagnosis. 3 Introduction In ACS patients, prompt diagnosis of acute coronary occlusion with early reperfusion therapy leads to a reduction in morbidity and mortality. However, initial 12-lead (12L) ECG has limited sensitivity (50-60%) for STEMI diagnosis [1] with over 20% of patients categorised as Non-STEMI on 12L ECG having an occluded artery in the early infarct setting [2]. Undiagnosed STEMI represents a high-risk patient group: at least 12.2% suffer death or nonfatal AMI at 6-months [3]. As such, rapid determination of an occluded coronary artery at presentation is paramount. BSPM ST-segment elevation (STE) improves sensitivity (76%) for AMI diagnosis over 12L ECG while retaining similar specificity (92%) [4]. Furthermore, in patients with chest pain and ST-segment depression only on initial 12L ECG, BSPM STE has sensitivity 91% and specificity 72% for AMI and acute coronary occlusion on angiography [5]. In the OCCULTMI trial, a large prospective multicentre study, it was shown that BSPM increased STEMI diagnosis by 27.5% over the standard 12L ECG [6]. In addition, patients in whom STE was only detected on BSPM had increased risk (OR 3.4) of death and recurrent AMI at 30-days post-presentation [6] due to a delayed revascularisation strategy. Epicardial Potentials (EPs) derived from Body Surface Potentials, a process known as inverse electrocardiography, further improves AMI diagnosis by removing the effect of the thoracic volume on electrical conduction [7]. However, to date the method used to derive EPs assumes that all patients have torso geometry similar to an athletic, adult male. In an increasingly obese population, individualised inverse electrocardiography has the potential to further improve AMI diagnosis, since an individualised torso geometry has shown potential in 4 deriving more accurate EPs than those calculated using an undiscriminating male standard [810]. The aim of this research was to assess if a patient-specific inverse solution would further improve the diagnostic sensitivity of BSPM for AMI. Methodology Phase 1 Torso Modelling Review of a hospital database of patients undergoing PET-CT scanning as part of their clinical care was undertaken (n=5450). Within each quartile (Q1 [BMI ≤ 20.0 kg/m2]; Q2 [BMI 20.1 - 30.0 kg/m2]; Q3 [BMI 30.1 - 40.0 kg/m2]; and Q4 [BMI > 40.0 kg/m2]), 12 patients were randomly selected (50% male) for analysis (n total = 48). PET-CT scanning was undertaken using a standard hospital protocol (supine position; from external auditory meatus to proximal femur), resulting in 260-300 regularly spaced crosssectional slices. Circumference, coronal/sagittal widths and cross-sectional area (CSA) of each plane were calculated [IDL version 7.0.0 (ITT Visual Information Solutions)]. Five coronal cross-sectional planes were chosen for comparison: Plane 1- sternal notch level; Plane 3 – nipple level; Plane 5 - umbilical level; Plane 2 - the cross-sectional plane equidistant between Planes 1 and 3; and Plane 4 - the cross-sectional plane equidistant between Planes 3 and 5. Following three-dimensional volume reconstruction, performed by 5 constructing a ‘stack’ of axial slices, values for torso volume, surface area and vertical displacement (distance between Planes 1 and 5) were obtained. Phase 2 Patient Population Between August 2009 and August 2011 we studied all patients admitted to our coronary care unit using either the emergency department or mobile coronary care unit. Those who fulfilled the following inclusion criteria were entered into the study: 1. Typical ischaemic-type chest discomfort of ≥ 20minutes duration, occurring at rest and presentation within 12 hours of symptom onset; 2. 12L ECG performed at first medical contact; 3. Informed consent to BSPM performed at first medical contact; and 4. Blood sampled for cardiac troponin T (cTnT) at 12hrs from symptom onset. Patients were excluded from analysis if they had: any condition precluding assessment of the ST-segment, i.e. left bundle branch block, right bundle branch block, left ventricular hypertrophy, concomitant digitalis therapy or ventricular pacing; received fibrinolysis, nitrates or glycoprotein IIb/IIIa inhibitor prior to initial 12L ECG or BSPM; or BSPM recorded >15mins after initial 12L ECG. In addition, demographic data and risk factors for coronary artery disease were documented. 6 Twelve-lead ECG analysis A 12L ECG was recorded at first medical contact (25mm/s and 10mm/mV). STE was measured at the J-point by a cardiologist who was blinded to all other clinical data. STE consistent with AMI (STEMI) was defined using the Minnesota code 9-2 [11] as ≥ 0.1mV STE in one or more of leads I, II, III, aVL, aVF, V5, V6 or ≥ 0.2mV STE in one or more of leads V1 – V4 [12, 13]. BSPM recording and analysis The BSPM was recorded with a flexible plastic anterior and posterior electrode harness and a portable recording unit (Heartscape Technologies, Inc.). The anterior harness contains 64 electrodes, including 3 proximal bipolar limb leads (Mason-Likar position) and a posterior harness with 16 electrodes. This lead configuration enables recording of 77 unipolar ECG signals with respect to the Wilson central terminal. During the interpretation process the electrodes are defined to represent anterior, lateral, inferior, high right anterior, right ventricular and posterior epicardial regions [7]. Harness application takes 3 – 4 minutes. BSPMs were recorded over 5 – 10 seconds at a sampling rate of 1 kHz and a bandwidth of 0.05 – 100 Hz and transferred into digital format for core laboratory analysis. The 80-lead BSPMs were uploaded and displayed on an IBM compatible computer running PRIME™ analysis software. All 80 leads were manually checked and those of unacceptable quality, i.e. where noise or movement artefact disallowed recognition of QRST variables, were marked and substituted using linear grid interpolation [7]. Any BSPM with >6 leads requiring interpolation were disregarded and these patients excluded from analysis. Printouts 7 were obtained from the processed BSPM of the 80-lead ECG and a colour-contour map displaying the amount of STE at the J point (ST0 isopotential map) and confirmed by a cardiologist blinded to the 12L ECG result. STE was measured from the ST0 point and defined by the following thresholds: anterior ≥ 0.2mV elevation; lateral/inferior/high right anterior/right ventricular ≥ 0.1mV elevation; posterior ≥ 0.05mV elevation; with infarctlocation described by the ST0 isopotential colour-contour map. The result of the PRIME™ diagnostic algorithm was noted. AMI diagnosis Diagnosis of AMI was made when cTnT ≥ 0.1µg/L [Roche Diagnostics, Switzerland]. Phase 3 Inverse Solution Using formulae derived from correlation analysis during Phase 1 (Figure 1), the dimensions (total volume and surface area; and sagittal/coronal widths, circumference and cross-sectional area at Planes 1-5) of a 352-node Dalhousie torso (Figure 2) were adjusted for each patient based on their individual BMI. BSPM data for each patient was exported from the PRIME software to obtain J-point values for each electrode. These were expanded using Laplacian interpolation to yield values at all points on an individualised 352-node Dalhousie torso [1417] (MATLAB R2009a [MathWorks Inc., USA]). Inverse calculations were performed using SCIRun (Version 4.5) software [18] and Tikhonov zero order regularisation to subsequently derive J-point potentials at 98 epicardial nodes (EPs) [18]. STE was defined as EP >0.3mV in any epicardial node [7, 19]. 8 Statistical analysis Data are presented as number (%) and mean ± standard deviation. Group comparisons were tested using the unpaired t test and χ2 test. Continuous clinical variables were tested by analysis of variance. A p-value < 0.05 was considered statistically significant. Correlations were made using Pearson’s correlation coefficient, with r2 ≥ 0.50 representing a strongly positive correlation. Sensitivity and specificity of the various diagnostic methods were calculated by comparing the prediction of AMI against cTnT ≥0.1µg/L. Statistical analysis was performed using SPSS version 17.0 for Windows (SPSS Inc, Chicago, Illinois). Ethical approval for BSPM was granted by the Local Ethics Committee. 9 Results Phase 1 Of the 48 patients studied (50% male; age 65 ± 16 yrs), results are shown in Table 1. Overall, no statistically significant differences were observed between male and female patients in any parameter. However, of those in Q1 (BMI ≤ 20.0 kg/m2 [n=12]), there were significant differences in both torso surface area and volume between male and female patients (p < 0.001). In addition, there were significant differences between circumferences in Planes 1-3 inclusive (p < 0.05) in this quartile. However, there were no significant differences between any measurements at any plane in either Q2 or Q3 (BMI 20.1 – 40.0 kg/m2). In Q4 (BMI > 40.0 kg/m2 [n=12]), all measurements showed no significant difference, except for both the coronal diameter in Plane 1 and the sagittal diameter in Plane 4 (p < 0.05). Analysis of correlation revealed a strongly positive relationship, i.e. r2 > 0.500, between BMI and all parameters measured at Planes 1-5 inclusive (Figure 1). Furthermore, BMI positively correlated with torso volume, i.e. r2 = 0.855 (Figure 1). Based on these linear relationships, formulae (y = ax +b, where x = BMI) were derived with which to individualise the Dalhousie torso in order to achieve a patient-specific inverse solution. Phases 2 and 3 Of the 400 patients who entered the study, 80 were excluded from analysis due to ECG confounders (n=72), a delay greater than15min between 12L ECG and BSPM (n=1) and therapy prior to BSPM (n=7). No patient was excluded for having >6leads of poor quality on 10 BSPM. Thus, 320 patients met the study criteria (57% male; age 62 ± 13 yrs) [Table 2], 180 (56%) of whom had AMI by cTnT definition. Patients with AMI had higher incidence of both hyperlipidaemia and cigarette smoking than patients without AMI (Table 2). However, nonAMI patients were more likely to have a positive family history of IHD and prior history of AMI and/or PCI than those patients with AMI. Otherwise, the baseline characteristics between groups were similar. Electrocardiographic diagnosis Overall, Minnesota code 9-2 STE was detected on initial 12L ECG in 132 (41%) patients (Table 3). Of these, 117 patients had AMI, i.e. cTnT ≥ 0.1µg/L (sensitivity 65%, specificity 89%). Overall, BSPM PRIME algorithm STE was detected in 160 (50%) patients, having sensitivity 81% and specificity 90% for AMI diagnosis (Table 3). Upon using a standard unadjusted Dalhousie torso for each patient, EPs derived by inverse electrocardiography improved both sensitivity and specificity for AMI diagnosis (Table 3). However, when the standard torso was individualised using BMI for each patient to derive EPs, sensitivity was further improved to 98% and a reduction in specificity was observed (79%). Of those with AMI by cTnT definition but without Minnesota STE on 12L ECG at presentation (n=63), 29 patients had STE detected by BSPM and 41 patients had STE detected when a standard torso was used for the inverse solution: sensitivity 65%, specificity 94% for AMI diagnosis. Of the remaining 22 patients, STE was detected in 18 patients when an individualised torso was used to derive EPs (Figure 3). 11 Discussion Current guidelines recommend a door-to-balloon time of ≤ 90mins for STEMI patients undergoing primary PCI [13, 20] with multiple studies having demonstrated increasing morbidity and mortality with treatment delay in these patients [20]. Therefore, patients with acute coronary occlusion, i.e. STEMI, not detected by 12L ECG are more likely to undergo a delayed revascularisation strategy and are thus at increased risk of an adverse prognosis. To date, the 12L ECG remains the initial bedside investigation in the triage of patients presenting with acute ischaemic-type chest pain at rest despite its poor sensitivity for AMI diagnosis – a study has confirmed that the 12L ECG STE has only 50% sensitivity for AMI diagnosis using contrast-enhanced cardiac magnetic resonance imaging as a gold-standard [21]. In patients with Non-STE AMI, current guidelines suggest a maximum delay to invasive strategy of 72h from admission in those without high-risk clinical features, i.e. haemodynamic instability, malignant arrhythmia and/or refractory ischaemia [22]. In addition, the Occluded Artery Trial has shown no prognostic benefit of PCI for patients with an infarct artery that is occluded ≥24 hours after symptom onset [2]. It is therefore of utmost importance that stable patients with acute chest pain, an acutely occluded coronary artery and STE ‘missed’ by the standard 12L ECG are identified at the earliest possible time-point in order to facilitate emergent revascularisation and improve outcomes. BSPM has been shown to have sensitivity 76%, specificity 92% for AMI diagnosis in consecutive patients presenting with acute ischaemic-type chest pain at rest (n=755) [4]. Improvement in sensitivity over the 12L ECG (sensitivity 68%) has been shown to be mainly due to detection of STE in the high right anterior, posterior, and right ventricular territories [4]. In the OCCULT-MI trial, BSPM increased STEMI detection by 27.5% over the 12L 12 ECG [6]. In this OCCULT-MI trial, STE on BSPM was associated with an increased risk of death and recurrent MI at 30-days (OR 3.4) in those without initial 12L ECG STEMI [6], due to a non-emergent revascularisation strategy. Previous work from our centre has shown improvement in AMI detection by an epicardial algorithm using a standard torso in patients with a non-diagnostic 12L ECG [7]. Compared to other electrocardiographic techniques assessed, an epicardial algorithm correctly identified 53/83 patients with a non-diagnostic 12L ECG at presentation as AMI (sensitivity 64%, specificity 99%) - findings comparable to those presented in this study (41/63 patients; sensitivity 65%, specificity 94%). Furthermore, in the 427 patients with acute ischaemic-chest pain studied by Owens et al [7], the 12L ECG in combination with an epicardial algorithm (adjusted for confounders) derived EPs that had sensitivity 85% and specificity 98% for AMI diagnosis - using standard CK/CKMB incremental elevation as the diagnostic standard [7]. Again, this is comparable to the results presented here, when a standard torso is used for the inverse calculation (sensitivity 88%, specificity 95%) [Table 3]. As shown in this study, there remain a significant proportion of patients with AMI, without Minnesota STE on 12L ECG and STE on either BSPM or EPs derived from an inverse solution using a standard torso (22/63 patients [35%]). In this study, the development of a novel method to ‘tailor’ the standard torso for the individual patient based on BMI, significantly improves the diagnostic sensitivity of epicardial potentials derived from the BSPM for AMI (sensitivity 98%) [Figure 3]. However, the detection of STE in patients without AMI when an individualised torso is used in the inverse calculation has resulted in a 13 reduction in specificity by 16% (Table 3). This may be in part explained by the lack of an individualised cardiac geometry used in our model. Further study is required in this regard, in addition to a prospective randomised-control study of our method in a real-world population, in order to fully assess prognostic impact. Study Limitations This is the first non-invasive study attempted in man to improve diagnosis of acute myocardial infarction using inverse electrocardiography (individual torso model) related to body habitus within the early hours of the onset of chest pain. Yes, there were limitations but the results are encouraging. Protocol development and patient recruitment were undertaken between 2009-11. Since this time: (a) the 12-lead ECG criteria used for STEMI diagnosis have been further refined; and (b) the cTnT assay used in this study to define AMI has been superseded by incremental rise in the high-sensitivity cTnT assay. Furthermore, this study was undertaken as part of a research protocol where coronary angiography was performed when clinically indicated. Future work Recently, in attempting to increase sensitivity for diagnosis of acute myocardial infarction early after the onset of symptoms the approach taken by other research workers has been at the expense of reduced specificity. Nevertheless, in an effort to improve the diagnostic specificity of the novel electrocardiographic method described in this work, effort should be made to individualise the cardiac geometry. In addition, BSPM and EP data obtained should 14 be correlated with angiographic findings. Prospective studies where the novel method described directly affects clinical decision-making should be undertaken, with long-term outcomes evaluated. Acknowledgements Financial support: Dr Michael J Daly is supported by The Heart Trust Fund (Royal Victoria Hospital), 9B Castle Street, Comber, Newtownards Northern Ireland BT23 5DY. 15 References 1. Owens CG, McClelland AJJ, Walsh SJ, et al. Prehospital 80-LAD mapping: Does it add significantly to the diagnosis of acute coronary syndromes? J Electrocardiol 2004;37suppl:223-32 2. Menon V, Ruzyllo W, Carvalho AC, et al. Infarct artery distribution and clinical outcomes in occluded artery trial subjects presenting with non-ST-segment elevation myocardial infarction (from the long-term follow-up of Occluded Artery Trial [OAT]). Am J Cardiol 2013;111:930-35 3. Collinson J, Flather MD, Fox KAA, et al for the PRAIS-UK Investigators. Clinical outcomes, risk stratification and practice patterns of unstable angina and myocardial infarction without ST elevation: Prospective Registry of Acute Ischaemic Syndromes in the UK (PRAIS-UK). Eur Heart J 2000;21:1450-57 4. Owens C, McClelland A, Walsh S, Smith B, Adgey J. Comparison of value of leads from body surface maps to 12-lead electrocardiogram for diagnosis of acute myocardial infarction. Am J Cardiol 2008;102:257-65 5. Daly MJ, Finlay DD, Guldenring D, et al. Detection of Acute Coronary Occlusion in patients with Acute Coronary Syndromes presenting with Isolated ST-segment Depression. Eur Heart J: Acute Cardiovasc Care 2012;1:128-135 6. Hoekstra JW, O'Neill BJ, Pride YB, et al. Acute detection of ST-elevation myocardial infarction missed on standard 12-lead ECG with a novel 80-lead real-time digital body surface map: primary results from the multicenter OCCULT MI trial. Ann Emerg Med 2009;54:779-88 16 7. Owens C, Navarro C, McClelland A, et al. Improved detection of acute myocardial infarction using a diagnostic algorithm based on calculated epicardial potentials. Int J Cardiol 2006;111:292-301 8. Huiskamp G, van Oosterom A. Tailored versus realistic geometry in the inverse problem of electrocardiography. IEEE Trans Biomed Eng 1989;36:827-835 9. Huiskamp GJM, van Oosterom A. Heart position and orientation in forward and inverse electrocardiography. Med Biol Eng Comput 1992;30:613-620 10. Hoekema R, Uijen GJH, van Erning L, van Oosterom A. Interindividual variability of multilead electrographic recordings. J Electrocardiol 1999;32:137-148 11. Menown IBA, Mackenzie G, Adgey AAJ. Optimizing the initial 12-lead electrocardiographic diagnosis of acute myocardial infarction. Eur Heart J 2000;21:275-283 12. Antman EM, Anbe DT, Armstrong PW, et al. ACC/AHA guidelines for the management of patients with ST-elevation myocardial infarction (committee to revise the 1999 guidelines). J Am Coll Cardiol 2004;44:671-719 13. Antman EM, Hand M, Armstrong PW, et al. 2007 focused update of the ACC/AHA 2004 guidelines for the management of patients with ST-elevation myocardial infarction: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Canadian Cardiovascular Society; American Academy of Family Physicians; American College of Cardiology; American Heart Association. J Am Coll Cardiol 2008;51:210-47 14. Oostendorp TF, van Oosterom A, Huiskamp G. Interpolation on a triangulated 3D surface. J Comput Phys 1989;80:331-43 17 15. Horacek BM, Clements JC. The inverse problem of electrocardiography: a solution in terms of single- and double-layer sources of the epicardial surface. Math Biosci 1997;144:119-54 16. Horacek BM, Warren JW, Penney CJ, et al. Optimal electrocardiographic leads for detecting acute myocardial ischaemia. J Electrocardiol 2001;34(Suppl):97-111 17. Horacek BM, Sapp JL, Penney CJ, Warren JW, Wang JJ. Comparison of epicardial potential maps derived from the 12-lead electrocardiograms with scintigraphic images during controlled myocardial ischaemia. J Electrocardiol 2011;44:707-12 18. SCIRun: A Scientific Computing Problem Solving Environment, Scientific Computing and Imaging Institute (SCI) 19. Navarro C, Owens C, Riddell J, et al. The use of calculated epicardial potentials improves significantly the sensitivity of a diagnostic algorithm in the detection of acute myocardial infarction. J Electrocardiol 2003;36(Suppl):127-32 20. Pride YB, Tung P, Mohanavelu S, et al for the TIMI Study Group. Angiographic and Clinical Outcomes among patients with acute coronary syndromes presenting with isolated anterior ST-segment Depression. J Am Coll Cardiol Intv 2010;3:806-11 21. Martin TN, Groenning BA, Murray HM, et al. ST-Segment deviation analysis of the admission 12-lead electrocardiogram as an aid to early diagnosis of acute myocardial infarction with a cardiac magnetic resonance imaging gold standard. J Am Coll Cardiol 2007;50:1021-8 22. Roffi M, Patrono C, Collet J-P, et al. for the Task Force for the Management of Acute Coronary Syndromes in Patients Presenting without Persistent ST-Segment Elevation of the European Society of Cardiology (ESC). 2015 ESC Guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation. Eur Heart J 2016;37:267-315 18 Figures legends: Figure 1: Axial CT slice at Plane 3 (nipple level) with torso reconstruction and Planes 1-5 shown. Graphical plots of torso volume (top right) and CSA at Plane 3 (nipple level) [bottom right] against BMI, with both r2 values and linear equations shown. BMI = body mass index; CSA = cross-sectional area; USN = upper sternal notch Figure 2: 352-node Dalhousie torso Figure 3: Case example BMI = body mass index; BSPM = body surface potential map; cTnT = cardiac troponin-T; ECG = 12-lead electrocardiogram; EP = epicardial potentials; STE = ST-segment elevation. 19