Skip to main content
eScholarship
Open Access Publications from the University of California

UC Berkeley

UC Berkeley Electronic Theses and Dissertations bannerUC Berkeley

Addressing Cardiac CEST Limitations for Clinical Implementation

Abstract

For decades, cardiovascular disease has been the leading cause of death globally, presenting a significant public health challenge. Cardiovascular disease encompasses a large range of disorders affecting blood vessels and the heart, including cardiomyopathies, stenosis, and atrial fibrillation. Current diagnostic technology, such as electrocardiograms, echocardiography, and magnetic resonance imaging, have improved the accuracy of cardiovascular disease diagnostics but often rely on symptomatic presentation and the presence of cardiac functional loss. The lack of quantitative, non-invasive, and holistic methods of identifying early indicators of cardiovascular disease, such as microvascular dysfunction, has resulted in late-stage diagnosis and delayed therapeutic treatment initiation. Early detection of these diseases could slow down the progression of cardiovascular disease, leading to improved patient outcomes and reduced mortality rates.

Chemical exchange saturation transfer (CEST) MRI is a non-invasive imaging technique that facilitates quantification of endogenous and exogenous contrasts. This technique enables the detection of metabolic content, on the order of 10 mM, without the need for radiation or external hardware. Ongoing CEST research significantly improved accuracy and addressed technical limitations, as shown by recent larger clinical use of CEST in neuroimaging for tumor identification and classifications. Similarly, studies with both small and large animal models, as well as human subjects, have demonstrated the efficacy of CEST in cardiac applications. Important cardiac CEST targets include fibrosis, energetic dysfunction, hypertrophy, and other indicators of cardiovascular disease. Despite its clinical potential, widespread use of cardiac CEST is hindered by various limiting factors, such as B-field inhomogeneities, scan time, and cardiac motion. To address these limitations, I propose three study strategies: (1) the use of tailored spatial-spectral saturation pulses to compensate for B1 inhomogeneity, (2) an overhaul of cardiac CEST scanning protocol to reduce CEST acquisition time, and (3) initiating a proof-of-concept cardiac study which compares the improved CEST protocol with a clinically validated diagnostic method.

First, spatial-spectral RF pulses were designed to compensate for B1 variations and tested in simulations, phantoms, and in vivo cardiac scans. B1-field maps were used to individually tailor an optimized saturation field around the region of interest. This study recruited ten healthy subjects who were subsequently scanned following a pulsed CEST protocol with frequency offsets rangingfrom -10.0 ppm to 10.0 ppm. In 2-pool simulations, the use of SPSP pulses improved the significantly improved spatial homogeneity around the myocardium. Following CEST acquisition and data analysis, spatial-spectral pulses resulted in reduced total creatine contrast variation across the myocardium from 10.0 ± 4.3% to 6.1 ± 3.5% (p < 0.05). Moreover, the use of spatial-spectral saturation resulted in a significant reduction of CEST contrast around 2 ppm. The same CEST protocol on a 50 mM creatine and 50 mM phosphocreatine phantom revealed phosphocreatine-specific contrast generation at physiological temperature with spatial-spectral saturation but combined phosphocreatine and creatine contrast generation with Gaussian saturation. The use of spatial-spectral saturation pulsed CEST enabled ratiometric mapping of phosphocreatine to total creatine CEST contrast in vivo at 3T.

Second, to validate the phosphocreatine selectivity in clinical applications, a set of studies was conducted to minimize the total acquisition time of cardiac CEST. In the first study, more than 50 images were acquired per subject, each with varying frequency offset, for a total scan time of approximately 60 minutes. However, reducing the number of frequency offsets to only those significant to creatine and phosphocreatine enabled sparse ratio analysis similar to the already established magnetization transfer ratio asymmetry. Through the use of cardiac (n = 10) and skeletal muscle (n = 14) scans and analysis, we correlated the sparse ratio method with long scan results and found comparable phosphocreatine to total creatine ratios. Furthermore, the initial spatial-spectral design method required offline analysis which was time and resource intensive. A cardiac study (n = 6) was conducted to compare CEST contrast generated with tailored and nontailored spatial-spectral pulses and found no significant spatial variation between the two saturation profiles, eliminating the need for offline tailoring design in the future. Through these studies, the total scan time for cardiac CEST was significantly reduced from 60 minutes to approximately 4 minutes and the scanning procedure was streamlined for clinical use.

Third, the shortened cardiac CEST scanning procedure was evaluated in a clinical setting, requiring initial sequence installation and testing followed by patient recruitment and imaging. Initial tests conducted on healthy subjects resulted in similar phosphocreatine to total creatine ratios to those observed in previous healthy populations. Patients suspected of having hypertrophic cardiomyopathy, myocardial fibrosis, or other symptomatic cardiovascular diseases were recruited for cardiac diagnostic imaging. Patients also underwent late gadolinium enhancement imaging for fibrotic tissue mapping which was subsequently compared to the CEST ratio mapping. The results from sparse cardiac CEST ratio maps of phosphocreatine and total creatine showed similar distributions to the late gadolinium enhancement images.

In conclusion, this work aimed to identify and mitigate barriers in clinical cardiac CEST. The studies aided in the development of a method for quantifying myocardial energetics noninvasively and quickly through mapping phosphocreatine and creatine in the myocardium. Initial clinical work provided guidance for future cardiac CEST imaging, which may be used for diagnosing early stages of heart-related cardiovascular diseases.