P5287Influence of contrast and spatial resolution on myocardial strain results using feature tracking MRI

Abstract

Abstract Background Feature tracking is a new post-processing algorithm for calculating myocardial strain from cardiac magnetic resonance (CMR) cine images. Quantification of myocardial deformation promises to detect myocardial abnormalities early in the course of disease. However, no uniform standards for image acquisition and processing are available. Purpose Analysing the influence of spatial resolution and contrast agent on quantitative myocardial strain results. Methods Seventy-five patients in sinus rhythm were prospectively enrolled and underwent a standardised CMR exam using a 1.5 Tesla scanner. For feature tracking analysis, steady-state free-precession (SSFP) cine images in long and short axes were acquired in 3 variants: 1. Native standard cine (spatial resolution 1.4x1.4x8mm3). 2. Native cine with lower spatial resolution (2.0x2.0x8mm3). 3. Standard cine equal to variant 1 but acquired 5 minutes after i.v.-administration of 0.2mmol/kg body weight gadoteracid. Group A consisted of 50 patients with normal left ventricular ejection fraction, no regional wall motion abnormality and no evidence of fibrosis based on Late Gadolinium Enhancement. Group B comprised 25 patients with evidence of infarct-like myocardial scar in the analyzed planes. Using dedicated feature tracking software, circumferential, longitudinal and radial strain were calculated as mean of the six midventricular myocardial segments for group A. For group B, the analysis focused on the infarcted segments. Results Figure 1 summarises the strain results, given as mean standard deviation. (The p-values stem from paired t-tests (group A) and Wilcoxon tests (group B), respectively. Significance is highlighted by *). In group A, circumferential and radial strain decreased significantly after Gadolinium administration, while longitudinal strain increased significantly with lower spatial resolution. In group B, circumferential, longitudinal and radial strain results of the infarcted segments increased significantly with lower spatial resolution, while Gadolinium had no statistically significant influence on strain results. Figure 1 Conclusion Quantitative myocardial strain results obtained by CMR feature tracking are significantly influenced by spatial resolution and Gadolinium administration. Standardised image acquisition seems fundamental for reliable and comparable CMR strain analysis.