Feasibility of Strain Encoded Magnetic Resonance (SENC) at 0.55T.

Abstract

Background: Low-field (<1.0T) wide-bore cardiovascular magnetic resonance (CMR) has the potential to improve accessibility by reducing costs and accommodating severely obese and claustrophobic patients. However, intrinsically reduced signal-to-noise ratio (SNR) may affect techniques such as strain-encoded magnetic resonance (SENC), a method to quantify regional strain that may be more sensitive than global function measurements to detect abnormalities. We sought to characterize global and segmental strain derived from SENC on a wide-bore, low-field system in healthy human subjects and a porcine model of myocardial infarction.

Study type: Original Research METHODS: A segmented k-space, spoiled gradient echo prototype SENC sequence was implemented on a 0.55T system with an 80cm bore. A dynamic phantom and sixteen healthy volunteers (mean age 31yrs, 10 female) were scanned at 0.55T and 1.5T. Ten of the subjects were scanned twice at each field strength to evaluate scan-rescan repeatability. In volunteers, t-tests were used to compare global strain results; global and segmental strain reproducibility between field strengths and scan-rescan repeatability were assessed via Bland-Altman analysis and intraclass correlation (ICC) methods. Additionally, adjunctive SENC followed by late-gadolinium enhancement (LGE) was acquired at 0.55T eight weeks post myocardial infarction (MI) in an ongoing study of a porcine model (n=6) of non-reperfused MI. Porcine left ventricular (LV) segments were categorized based on LGE and compared to resultant segmental strain via one-way ANOVA.

Results: Mean phantom strain showed no significant differences between field strengths (p > 0.10). In volunteers mean LV global longitudinal (GLS) and circumferential strain (GCS) were -19.4% ±1.1 and -20.4% ±0.9 at 0.55T compared to -18.7 ±1.4% and -19.2% ±1.6 at 1.5T (p>0.10). For both 1.5T vs 0.55T reproducibility and scan-rescan repeatability, LS proved to have better agreement than CS, and mean biases were low for both global and segmental comparisons throughout. Limits of agreement were good for global strain comparisons, but were notably wider when comparing segmental values, especially circumferential strain reproducibility and 0.55T scan-rescan repeatability. ICC analysis of pooled LV segmental strain showed good LS agreement between and within field strengths (0.78-0.89), but was fair for CS between 1.5T vs 0.55T (0.60) and CS 0.55T repeatability (0.64). In the pigs, LGE demonstrated an expected territory of infarction; segmental LS in LGE+ vs remote segments was -10.8% ±4.0 vs -16.8% ±5.1; p<0.001. Segmental CS in LGE+ vs remote segments was -11.9% ±2.7 vs -14.6% ±2.7; p=0.0011.

Conclusions: Our results support the feasibility of SENC at 0.55T, with accurate phantom measurements, good agreement of global values in human volunteers, and correlates of functional impairment with known MI territory. Reproducibility between field strengths showed minimal systemic bias but at times substantial limits of agreement. Repeatability of global and segmental longitudinal strain at 0.55T was similar to established 1.5T performance, although circumferential strain was notably poorer. LV circumferential strain may lack sufficient reliability in its current implementation for use at 0.55T.