Document resource
Introduction Efficient electromechanical coupling is fundamental to normal cardiac function, and its disruption is a feature of cardiomyopathy and myocardial ageing. However, conventional tools assess electrical and mechanical function separately: 12-lead electrocardiography lacks regional resolution, while myocardial strain detects dysfunction without electrical context. Cardiovascular magnetic resonance–guided ECG imaging (CMR-ECGI) enables non-invasive, high-resolution mapping of cardiac activation and repolarisation, allowing assessment of regional electromechanical coupling. We hypothesised that ECGI-derived repolarisation abnormalities would be associated with impaired left ventricular (LV) myocardial strain, independent of hypertrophy and fibrosis.Methods Participants from MyoFit46, a prospectively recruited age-homogeneous cohort, underwent 5-minute 206-lead ECGI using a CMR-ECGI vest, followed by 3-Tesla CMR including cines, T1 mapping, and late gadolinium enhancement ( figure 1). Averaged body-surface ECGI signals were combined with CMR-derived heart–torso geometry to reconstruct epicardial unipolar electrograms and isochrone maps using an inverse solution developed at St Louis University.Global and 16-segment LV ECGI metrics included QRS and T wave amplitude, activation time (AT), repolarisation time (RT), and activation recovery interval (ARI), with heart-rate correction where appropriate (RTc, ARIc). Spatial dispersion was defined as the within-subject range (maximum–minimum), and spatial gradients as the mean parameter change per inter-electrogram distance. Global and segmental longitudinal, circumferential, and radial strain (GLS, GCS, GRS) were quantified by CMR feature-tracking. Associations between global ECGI metrics and strain were assessed using multivariable linear regression adjusted for sex, indexed LV mass, and extracellular volume. Segmental analyses used linear mixed-effects models with the same adjustments.Results 409 participants were analysed (mean 76.7 years; 55.5% female). Globally, prolonged RTc was independently associated with reduced GLS, GCS, and GRS ( table 1, figure 2A). Increased ARIc was associated with reduced GLS and GRS, while higher ECGI-derived QRS and T-wave amplitudes were associated with increased GRS. Greater dispersion of RT, RTc, ARI, and ARIc was associated with reduced GRS.Segmentally, prolonged ARIc was associated with impaired GRS, whereas higher RT dispersion and RT/RTc gradients were associated with impaired GLS (table 2, figure 2B).Conclusion In older adults without overt cardiomyopathy, ECGI-derived repolarisation abnormalities were associated with impaired global and regional LV strain, independent of hypertrophy and fibrosis, indicating subclinical electromechanical uncoupling. By bridging this electro-structural gap, CMR-ECGI may aid identification of subclinical phenotypes or enable tracking of disease progression.Abstract 472 Figure 1Workflow for ECGI-CMR acquisition. 5-minute ECGI recording followed by 3T CMR. ECGI data combined with heart-torso geometry to reconstruct epicardial electrograms and maps. Strain derived from cine imaging. Abbreviations: CMR, cardiac magnetic resonance, LV, left ventricle, RV, right ventricle. Others as per Table 1.Abstract 472 Figure 2Forest plot of significant (A)global and (B) segmental ECGI-strain associations. Effect size and 95% confidence interval represents change in strain per unit change in ECGI variable (mV for T wave amplitude, ms for others). Abbreviations as per Table 1.Abstract 472 Tabel 1Association between global ECGI and strain metrics Strain ~ ECGI metricβTiming featuresGLS ~ RTc-0.038 (-0.063, -0.013), 0.003GCS ~ RTc-0.047 (-0.081, -0.013), 0.006GRS ~ RTc-0.173 (-0.294, -0.051), 0.006GLS ~ ARIc-0.035 (-0.060, -0.009), 0.008GRS ~ ARIc-0.142 (-0.265, -0.019), 0.024Waveform featuresGRS ~ QRS amplitude6.682 (-0.973, 12.391), 0.022GRS ~ T wave amplitude19.732 (2.535, 36.928), 0.020Dysynchrony featuresGRS ~ RT dispersion-0.304 (-0.519, -0.090), 0.006GRS ~ RTc dispersion-0.308 (-0.512, -0.103), 0.003GRS ~ ARI dispersion-0.271 (-0.463, -0.079), 0.006GRS ~ ARIc dispersion-0.273 (-0.458, -0.088), 0.004Values presented as effect size (95% confidence interval), p value per 1 unit change.Abbreviations: ARI, activation repolarisation interval, ARIc, corrected activation repolarisation time, ECGI, electrocardiographic imaging, GCS, global circumferential strain; GLS, global longitudinal strain; GRS, global radial strain, RT, repolarisation time, RTc, corrected repolarisation time.Abstract 472 Tabel 2Association between segmental ECGI and strain metrics Strain ~ ECGI metricβTiming featuresGRS ~ ARIc-0.062 (-0.122, -0.003), 0.040Dysynchrony featuresGLS ~ RT dispersion0.016 (0.001, 0.030), 0.035GLS ~ RT gradient0.530 (0.240, 0.061), 0.027GLS ~ RTc gradient0.452 (0.004, 0.899), 0.048Values presented as effect size (95% confidence interval), p value per 1 unit change. Abbreviations as per Table 1.