NOVEL MRI APPROACHES FOR QUANTITATIVE ASSESSMENT OF CORONARY ARTERIES AND VENTRICULAR FUNCTION
Embargo until
Date
2014-06-02
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Johns Hopkins University
Abstract
Despite advances in screening techniques and prevention guidelines, cardiovascular disease remains the leading non-communicable disease with 17 million deaths worldwide in 2008. Approximately 6% of American adults had a history of coronary artery disease in 2010, and there were more than 700,000 new cases of myocardial infarction in the US in the same year. Imaging technologies have played an essential role in the diagnosis of cardiovascular diseases. Cardiovascular magnetic resonance imaging (CMR), noninvasive in nature and providing many of the markers of cardiovascular disease, has gained much popularity in the past few decades. However, restricted spatial and temporal resolution, limited tissue contrast, long acquisition, and complex post-processing requirements are some of the main challenges that hinder mainstream use of CMR in practice. The increasing availability of high field scanners and their enhanced signal, which can be traded for faster acquisitions, larger volumetric coverage, and improved resolution have introduced new frontiers in CMR. This thesis focuses on two major applications of this technology: coronary artery angiography and ventricular function imaging. Angiography with CMR is especially challenging due to the small caliber of coronary arteries, their tortuosity, their wide range motion, and their proximity to myocardial and fat tissues.
The first contribution of this thesis was the design, development, and validation of a refined T2Prep sequence to improve coronary artery conspicuity by deliberately taking advantage of the coronary blood flow. The technique was designed and developed on a 3.0T scanner, where the user could freely adjust angulation and position of a three-dimensional (3D) slab to avoid covering the ventricular blood and saturating the inflowing spins. Results in 22 in vivo healthy coronary artery segments demonstrated a 35% increase in signal to noise ratio (42.3 ± 2.9 versus 31.4 ± 2.2, P < 0.0001) and a 34% increase in contrast to noise ratio (18.6 ± 1.5 versus 13.9 ± 1.2, P = 0.009) between the proposed technique and the conventional T2Prep, respectively.
The second contribution was the development and validation of a robust accelerated bSSFP technique that enables imaging of the major coronary arteries in a single-breathhold at 3.0 T. To accelerate the acquisition, variable-rate selective excitation pulses were designed for a double oblique geometry and implemented on a 3.0 T scanner. Localized image-based shimming, based on acquired B0- and B1+-maps, was performed to minimize high field susceptibility-related artifacts. Fifteen healthy volunteers and three patients with coronary artery disease underwent examination, and acquisitions were repeated in nine subjects. The average vessel length of 100.5 ± 6.3 mm and sharpness of 55 ± 2% compared favorably with earlier reported implementations of bSSFP and gradient echo at 3.0 T. Measurements demonstrated a highly statistically significant degree of inter- and intra-observer, and inter-scan concordance.
Coronary arteries usually exhibit high variability in size and curvature and are often surrounded by other structures, which make automatic segmentation a challenging task. Quantitative analysis of coronary angiography images with CMR typically requires several manual interactions. The third contribution of this thesis was development of a modular approach based on the level set methods to track the vessel centerline, segment the vessel boundaries, and measure transversal area. Coronary magnetic resonance angiography was performed in ten healthy volunteers and cross-sectional area was measured throughout each vessel. A close agreement with R = 0.923 was demonstrated between area measurements from the proposed method and the widely used semi automatic Soap Bubble tool, while the new technique required only two user-selected points in each artery.
Coronary artery disease if left untreated leads to myocardial infarction and loss of ventricular function. CMR is the gold standard for the assessment of such adverse events. Despite recent advancements in 3D acquisitions, CMR function imaging is limited to one-dimensional (1D) or two-dimensional (2D), due to the lengthy and complex acquisition and analysis requirements for 3D techniques as well as an undocumented benefit for 3D information. The fourth contribution of this thesis was a complete assessment of myocardial deformation using bSSFP cine, zHARP tagging, and LGE infarct. The diagnostic accuracy of 3D strain analysis with 1D and 2D analyses in identifying the infarct and its adjacent regions from healthy myocardium was investigated in twenty farm pigs with surgically induced myocardial infarction. In summary, cumulative 3D information (circumferential, radial, and longitudinal strains) provided incremental diagnostic accuracy in delineating the nonviable myocardium in comparison with the 1D (circumferential shortening) or 2D (circumferential shortening and radial thickening) quantification of strain. In identification of adjacent segments, receiver operating characteristic analysis using the 3D strain multivariate model demonstrated a significant improvement compared to 1D and 2D models.
Description
Keywords
MRI, noninvasive coronary angiography, cardiac function imaging, cardiovascular disease