The human brain, one of the most complex structures known, is composed of more than 100 billion neurons that process, disseminate, transform and attract information through more than a 100 trillion synapses. Interactions among brain cells give us the freedom to think, feel, move and maintain homeostasis all at the same time. To understand the systematic communication among brain cells, we require not only knowledge at elementary levels, but also at macroscopic level - aimed at the discovery of the emerging patterns and properties of neuronal interactions. Here, we used diffusion imaging to reveal the organization of neural pathways by capturing subtle changes in white matter make-up through measures sensitive to fiber integrity and microstructure - otherwise not detectable with standard MRI techniques. In addition, tractography was performed to infer neural pathways and connectivity patterns, yielding additional, more complex mathematical metrics describing the connectomics of brain networks. To assess the brain's network, graph theory was used - a branch of mathematics employed to model the topological organization of the white matter structure. Similarly, algebraic connectivity was also applied, not previously seen in the context of brain networks, which uses linear algebra and matrix theory to study the properties of graphs. These methods have all contributed to the discovery of potential biomarkers that can aid the understanding of white matter deterioration in the brain; special focus was directed towards neurodegenerative diseases such as Alzheimer's disease and all of its clinical stages, as well as frontotemporal dementia.

Purpose

Methods

Results

Conclusion

Cardiac magnetic resonance imaging (MRI) is a proven technique for the evaluation of myocardial structure and function. An introduction to cardiac physiology is presented in Chapter 1, wherein a measure of LV rotational mechanics called LV twist is defined. LV twist is the apparent rotation of the LV apex relative to the LV base and provides insight into LV function beyond that traditionally reported in measures like ejection fraction. Importantly, cardiac MRI techniques are very well suited to evaluating cardiac structure and function and these methods are outlined in Chapter 2. In particular, MRI tagging can be used to non-invasively generate myocardial tissue landmarks that facilitate the qualitative and quantitative assessment of left ventricular (LV) myocardial deformation in both research and clinical settings. This thesis focuses on several developments. First, we developed a new tagging technique called complementary radial tagging (CRT) that generates a tagging pattern that better matches the annular shaped LV myocardium in the short-axis view. CRT also has better tag contrast during late diastolic phases and can be used to accurately measure the LV mechanics during all cardiac phases (Chapter 3).

In parallel with the development of the CRT technique we used conventional clinical tagging protocols to measure LV rotational mechanics quantitatively. Myocardial fibrosis is known to frequently occur in patients with Duchenne Muscular Dystrophy (DMD). The consequences of myocardial fibrosis on LV mechanics, however, are incompletely understood. In Chapter 4 we report on the LV rotational mechanics in patients with DMD with or without myocardial fibrosis. The results suggested that both DMD and the presence of myocardial fibrosis independently worsen LV rotational mechanics.

In Chapter 5 we report on the effects of conventional cardiac MRI exams, which require repeated breath holding and place a burden on some clinical patients, on measure of LV rotational mechanics. This breath hold paradigm presents two problems for patients with DMD who typically develop progressive respiratory impairment and the signs and symptoms of cardiac involvement at an early age. Currently, clinical protocols use both breath-hold and free breathing techniques, as needed, to acquire data. Chapter 5 compared the LV rotational mechanics between these techniques in healthy subjects and patients with DMD. It was found that free-breathing significantly decreases estimates of LV twist compared to breath hold measures. The results suggest that when using quantitative imaging biomarkers of LV rotational mechanics to monitoring disease progression or the response to therapy, especially in patients with DMD for whom decline in respiratory function is certain, it is important to use a free-breathing strategy for all studies to facilitate intra-subject longitudinal comparisons.

In Chapter 6, LV rotational mechanics were further evaluated in patients with mitral regurgitation. In patients with mitral regurgitation LV twist decreases, while CL-shear angle (an alternate measure of LV rotational mechanics) pseudo-normalized due to subtle changes in the heart’s geometry.

Lastly, in Chapter 7, a technique for acquiring two slices simultaneously – Controlled Aliasing In Parallel Imaging Results In Higher Acceleration (CAIPIRINHA) was implemented in a cardiac MRI tagging sequence. The application of CAIPIRINHA to a LV tagging sequence is shown to achieve similar estimates of peak LV twist in a single breath hold, which simplifies the exam and avoids measurement differences that may arise from data acquired in different breath holds.

In conclusion, this thesis reports on several technical developments and clinical applications related to estimating quantitatively the function of the left ventricle. Taken together these developments can be combined to provide fast and accurate estimates of LV rotational mechanics that provide insight to LV function beyond traditional measures.

Cardiac cine imaging for conscious pediatric patients and patients with respiratory and/or cardiac dysfunction is challenging due to the need for repeated, reproducible, and sometimes prolonged breath holds. Hence a free-breathing cardiac cine exam is essential for improved patient comfort. Cardiac cine exams are routinely performed using balanced steady-state free precession (bSSFP) imaging due to its high signal-to-noise ratio (SNR) efficiency. However, bSSFP imaging requires the use of a high flip angle for increased signal and a short repetition time to reduce image artifacts. These two requirements result in increased heat energy deposition, measured in specific absorption rate (SAR), in units of Watts per unit mass of the patient's weight (W/kg). FDA regulates SAR in order to ensure patient safety during MRI exams. As such the increased SAR that accords with using bSSFP can limit its application at higher field strengths (≥3T).

High SAR also poses the highest risk for RF induced lead-tip heating and potential tissue damage and may inhibit pacing and/or ICD shocks in patients with implanted pacemakers or ICDs. Hence, performing cardiac cine imaging with lower SAR bSSFP is essential for safe imaging of patients with devices as well as to minimize patient discomfort.

3D T2 weighted spin echo imaging is another SAR intensive pulse sequence due to the use of large flip angles. Extending the acquisition duration can reduce the sequence's SAR. As a consequence, 3D T2 weighted prostate imaging takes ~7 minutes to acquire, which leads to patient discomfort as well as increased sensitivity to motion artifacts.

In this dissertation, variable flip angle (VFA) bSSFP techniques have been developed to lower the SAR or improve the signal and contrast of bSSFP for free-breathing cardiac cine applications, 3D T2 weighted prostate imaging and non-contrast MR arteriography of the lower leg all with the aim of reducing patient discomfort while maintaining image quality.

The first objective was to determine the optimal bSSFP FA for cardiac cine imaging. Detailed Bloch simulations of stationary myocardium, flowing blood in the presence of imperfect slice profile and off-resonance effects were simulated and compared with in vivo imaging experiments of ten healthy subjects and seven patients with a range of ejection fraction and heart rate. The in vivo imaging in healthy subjects and clinical patients show that high blood-myocardium contrast can be obtained with a FA ~105°, when imaging a plane with predominantly through-plane flow such as the short-axis plane. However, if through-plane flow is limited, as may occur for patients with low ejection fraction or low heart rates or four-chamber and three-chamber imaging planes, then the FA should be limited to ~75°.

The second objective was to develop a low SAR 2D breath-hold, cardiac cine imaging technique with blood-myocardium contrast similar to conventional bSSFP imaging using VFA scheme, which was implemented by using an asynchronous k-space acquisition. Bloch simulations were performed to determine if multiple VFA acquisitions of stationary myocardium and flowing blood produces a steady state signal between acquisitions and was verified by phantom experiments. In vivo experiments were also performed in ten healthy subjects with several VFA-bSSFP schemes, and were compared to the conventional segmented bSSFP acquisitions. The novel asynchronous VFA-bSSFP technique can lower the SAR by at least 36% with similar blood-myocardium CNR compared to the conventional bSSFP cardiac cine imaging. Asynchronous VFA-bSSFP can also be used to increase the CNR by at least 28%, with similar SAR compared to conventional bSSFP.

The third objective was to apply the asynchronous VFA-bSSFP cardiac cine imaging technique for low SAR, 2D free-breathing, bellows gated acquisitions at 3T. Bloch simulations of stationary myocardium and flowing blood were performed to determine the optimal VFA scheme that maximizes the SAR reduction while maintaining blood-myocardium contrast similar to conventional bSSFP imaging. In vivo experiments were performed in ten healthy subjects in both short-axis and long-axis view and the image quality of the bellows gated cardiac cine images (FB-VFA) were compared to standard breath-hold imaging (BH-CFA) using both quantitative metrics and evaluation by two radiological experts. The SAR was reduced by 25% in FB-VFA compared to BH-CFA with similar blood-myocardium contrast and with image quality sufficient to perform global and regional cardiac function analysis.

The fourth objective was to develop a fast, 3D T2 weighted VFA-bSSFP technique (3D T2-TIDE) for prostate imaging at 3T. Conventional 3D fast spin echo techniques take longer acquisition duration for T2 weighted imaging due to the long recovery time between multi-shot acquisitions. This was overcome in 3D T2-TIDE imaging by: 1) designing a VFA scheme for lowering the SAR and hence improving the signal for imaging at 3T; 2) using the transient bSSFP signal for T2 weighted imaging; and 3) modifying the acquisition of the 3D ky-kz plane using spiral-out phase encode ordering for efficient and faster filling of the 3D k-space. Bloch simulations were performed to evaluate the T2 weighting, contrast between normal prostate tissue and tumor, and the point-spread function of the 3D T2-TIDE sequence. In vivo experiments were performed in ten healthy subjects. 3D T2-TIDE images were acquired in 2:54 minutes compared to 7:02 minutes for 3D FSE with identical imaging parameters with SNR efficiency that exceeds 3D FSE.

The fifth objective was to develop a VFA scheme to maintain constant transverse magnetization of bSSFP without RF phase alternation (VUSEnoalt). This novel VUSEnoalt sequence has high SNR transient signal but is very off-resonance sensitive. This property of VUSEnoalt sequence is used to differentiate the arterial and the venous blood vessels due to their inherent off-resonance differences and was evaluated for non-contrast enhanced MR arteriography of the lower leg. The VFA scheme of VUSEnoalt was determined using Bloch simulations and their off-resonance effects were evaluated. In vivo imaging experiments were performed in three healthy subjects and the SNR of arterial and venous blood was compared between VUSEnoalt, VUSE and conventional bSSFP. VUSEnoalt reduces the venous blood signal by a factor of ~3 compared to the arterial blood with reduced fat signal.

Overall, VFA bSSFP techniques can be used to lower the SAR or improve the signal for 2D free-breathing cardiac cine imaging, 3D T2 weighted prostate imaging, and non-contrast MR arteriography of the lower leg.

The complex, but highly organized arrangement of myocytes within the left ventricle (LV) underlies the anisotropic mechanical and electrical function of the heart. Whole heart surveys of microstructural remodeling are critical to our basic science understanding of cardiac pathophysiology. Recently, diffusion tensor magnetic resonance imaging (DT-MRI) has emerged as a means for quantifying global and regional microstructural remodeling.

Previous observations of myocardial microstructure have predominantly relied on histological methods, which are tedious and difficult to spatially register in three-dimensions (3D). DT-MRI enables the 3D evaluation of whole heart microstructure. Each diffusion tensor can be decomposed into a system of eigenvectors and eigenvalues. The eigenvectors of the diffusion tensor contain the diffusion orientation information within the tissue microstructure and have established correspondence with the myofiber (primary eigenvector) and myolaminar sheet orientations (secondary and tertiary eigenvectors). Meanwhile, the eigenvalues quantify the magnitude of diffusion along each eigenvector and characterize the overall shape of diffusion. DT invariants (e.g. trace, fractional anisotropy, or mode), which are functions of the eigenvalues alone, saliently characterize specific attributes of tensor shape and can be used to evaluate specific attributes of microstructural remodeling.

Chapter 1 of this dissertation provides an understanding of myocardial microstructure in health and disease. Chapter 2 provides an understanding of DT-MRI, and its capabilities in evaluating soft tissue microstructure. In Chapter 3 we used an ovine model to establish the correspondence between both the secondary and tertiary eigenvectors of the local diffusion tensor and the dual sheet orientations observed in local histology. In Chapter 4 we quantified microstructural remodeling in a porcine model of myocardial infarction and its border zone using DT-MRI and revealed that the border zone significantly remodeled compared to both infarct and remote myocardium, providing a unique microstructural environment. In Chapter 5 we revealed significant microstructural remodeling in a rabbit mode of pacing induced heart failure using DT-MRI within transmural regions as well as transmural heterogeneity of microstructure within both normal and heart failure myocardium. Chapters 6 and 7 describe the conclusions of this work and provide an outlook for continued research based on this work's achievements.

Diffusion weighted imaging (DWI) is a powerful quantitative magnetic resonance imaging (MRI) technique that can probe tissues in vivo at the microscopic level and provide insight into cellular microstructural environment. Cardiac DWI has great potential value in its ability to answer open questions regarding myocardial structure, dynamics, and remodeling. Unfortunately, several technical limitations of current DWI techniques make its application in the beating heart very challenging, which leads to erroneous or inconsistent results. Amongst the challenges are an extreme sensitivity to bulk physiological motion, low signal to noise ratios (SNR), long scan times, and geometric image distortions. In this dissertation, these limitations are addressed with novel technical developments applied to the DWI pulse sequence including convex optimized diffusion gradient waveform design and multi-parametric tissue characterization.

A brief introduction to Nuclear Magnetic Resonance (NMR) and MRI is provided in Chapter 1. This leads into a description of the fundamental components of a DWI acquisition in Chapter 2 and an overview of the current state of cardiac DWI in Chapter 3.

In Chapter 4, a novel DWI strategy called Convex Optimized Diffusion Encoding (CODE) is described. CODE is a mathematical framework that formulates diffusion encoding gradient design as a convex optimization problem and automatically generates motion compensated (MOCO) waveforms that achieve the shortest possible echo times (TE) and thus improve SNR. First and second order moment nulled CODE (CODE-M1M2) permits DWI that is robust to cardiac motion with higher SNR than an existing MOCO technique. First order motion compensated CODE-M1 also improves robustness to cardiac induced motion in liver DWI with higher SNR than M1 nulled bipolar DWI. CODE can also be used for non-motion compensated DWI and improves SNR compared with traditional monopolar DWI in the brain.

In Chapter 5 we present a multi-parametric DWI strategy that simultaneously yields maps of the apparent diffusion coefficient (ADC) and T2 relaxation time constant in the heart (T2+ADC). Typically, DWI protocols include multiple acquisitions with a range of diffusion encoding strengths (b-value), but with constant TE to isolate the effect of diffusion of the signal. The joint T2+ADC approach varies both b-value and TE within the acquisition to facilitate estimation of both ADC and T2 relaxation. T2+ADC permits joint reconstruction with no increase in scan time compared with DWI alone and no effect on ADC measurement.

In Chapter 6 we use CODE-M1M2 diffusion encoding to perform cardiac diffusion tensor imaging (cDTI) and generate maps of myocardial microstructure in healthy volunteers. cDTI can be used to map myocardial fiber and myolaminar sheetlet orientations, which can contribute to our understanding of ventricular microstructure in health and disease and facilitate sophisticated mechanical models of cardiac dynamics. However, it is important to understand the uncertainty underlying these measurements to inform interpretation and define acquisition limitations. We apply a previously described bootstrap technique to measure the uncertainty in the diffusion tensors derived from CODE-M1M2 cDTI and establish achievable levels of precision in clinically feasible scan times.

In Chapter 7 the CODE framework is extended to compensate for the effect of eddy currents, which are a common cause of image distortions in DWI and DTI. Diffusion encoding gradients must be very strong to encode microscopic molecular displacements and these strong gradient pulses induce unwanted eddy currents in conductive MRI hardware components. If not addressed, eddy currents lead to distorted images and corrupted diffusion parameter estimates. We incorporate an eddy current model into the CODE optimization framework to develop eddy current nulled CODE (EN-CODE). EN-CODE accomplishes eddy current nulling with TEs that are comparable to traditional monopolar encoding and much shorter than the established twice refocused spin echo (TRSE) technique for eddy current nulling.

The developments described in this dissertation represent an improvement in the flexibility, efficiency, and robustness of diffusion encoding. The CODE framework can also be easily modified to address additional constraints and thus may prove useful in currently unforeseen applications.

Purpose

Methods

Results

Conclusion

In this dissertation a novel method for rapid quantification of left ventricular (LV) twist was developed, validated, and clinically evaluated. LV twist is defined as the difference in rotation of the apex relative to the base of the heart. Alteration in LV twist has been shown the correlate with LV dysfunction. Rotation was measured using Fourier Analysis of STimulated echoes (FAST), an image acquisition and rapid processing method. FAST can be applied to cardiovascular magnetic resonance (CMR) images with myocardial tagging, which imparts a sinusoidal pattern on the myocardial tissue that deforms with the LV during contraction and relaxation. First, FAST estimates of LV twist were validated using intra- and inter-observer comparison, intra- and inter-scan comparison, and comparison to values derived from a `gold-standard' technique. Second, FAST was used in conjunction with a CMR imaging technique designed to lengthen the duration of myocardial tag detection to acquire diastolic twist information. Third, FAST was applied to an imaging technique developed to minimize chemical shift induced myocardial tagging pattern artifacts, which produces more accurate measurements of LV twist. Fourth, FAST was used in combination with two free-breathing techniques to determine the impact of free-breathing on FAST estimates of LV twist. Fifth, FAST was used to characterize differences in LV twist in patients with moderate and severe mitral regurgitation. LV twist measurements derived from FAST showed good agreement for the intra- and inter-observer study and the intra- and inter-scan study. Additionally, systolic LV twist values matched well with the `gold-standard'. When used in combination with the improved myocardial tag duration, FAST estimates of diastolic LV twist matched well with literature values. LV twist derived from the normal imaging technique compared with the artifact minimizing technique did not detect any significant differences, however this study was performed in healthy subjects, specific patient populations may provide different results. The free-breathing study detected a significant decrease peak LV twist due to free-breathing compared with breath-held imaging derived values. A significant decrease in peak LV twist was detected in patients with moderate and severe mitral regurgitation. Overall, the FAST method for quantitative LV twist measurement has been validated and shown to produce accurate measurements of twist with limited user interaction.

Cardiovascular diseases are the leading cause of death globally and atherosclerosis is the primary underlying condition. Magnetic resonance imaging (MRI) is a promising diagnostic tool for evaluating atherosclerosis because of its excellent tissue contrast non-invasiveness. However many vascular MRI techniques have not been utilized clinically due to practical issues related to imaging speed, workflow, contrast media, and image quality. The primary focus of the work in this dissertation is to address some of these limitations and improve the capability of vascular MRI to accurately detect and characterize atherosclerotic lesions. Specifically, there are four areas of improvements of vascular MRI:

First, an online respiratory self-gating method (ADIOS) was developed for noncontrast magnetic resonance angiography of renal arteries. Some drawbacks of the conventional navigator gating method, including complicated planning and saturation artifacts were eliminated. Testing on healthy subjects confirmed that the proposed method was able to provide high-quality visualization of renal arteries with no navigator-induced artifacts, simplified setup, and shorter scan time.

Second, a diffusion-prepared turbo-spin-echo technique (DP-TSE) was developed to enable 3D high resolution diffusion weighted imaging for carotid plaque characterization without the need of contrast agent. The proposed method showed significantly higher vessel wall visibility, less distortion, and less partial volume effect. Apparent diffusion coefficient demonstrated the potential to differentiate lipid core from fibrous plaque tissue.

Third, a black-blood imaging technique (DANTE-SPACE) was developed to improve flow signal suppression in vessel wall imaging. A combined carotid and intracranial arterial wall imaging protocol was designed and tested. DANTE-SPACE demonstrated significantly improved arterial and venous blood suppression compared with conventional methods. Preliminary clinical study showed that the proposed method had the potential to detect vessel wall dissection and accurately differentiate plaque components.

Fourth, an imaging method for coronary plaque characterization was developed to allow simultaneous dark-blood T1-weighted imaging and bright-blood imaging. An image-based affine motion correction algorithm was designed to allow 100% respiratory gating efficiency. Results from healthy subjects demonstrated the feasibility of whole-heart coronary plaque imaging with isotropic high resolution. Preliminary results from patients with coronary artery disease showed the potential of this technique to detect intra-plaque hemorrhage and inflammation in the coronary wall.

MRI examinations for patients with active implantable medical devices (AIMDs) pose significant safety concerns. Among them, induced currents by the radiofrequency (RF) electromagnetic field can result in tissue damage. While the magnetic component of the RF field (B1) is necessary for spin excitation, the electric component does not have a significant role during an MRI exam but it is mostly responsible of the induction of currents that get concentrated by the AIMD. The induced current propagates in the AIMD lead and when it reaches the end of the lead, a temperature increase is observed at the lead,

ii

commonly known as lead-tio heating (LTH), which can damage the surrounding tissue. A large number of factors impact the magnitude of the induced currents. Therefore, the analysis can be extremely complex, which contributes to difficulty associated with understanding how to safely examine patients with an AIMD using MRI. Consequently, and unfortunately, MRI examinations for patients with AIMDs are often denied. To reduce safety risks, device manufacturers have created MR-conditional AIMDs which under specific conditions safety risks are reduced. The conditions often include the specific absorption rate (SAR). SAR is a measure to estimate RF exposure measured by the MRI scanner which is often limited to a certain value to mitigate temperature increase. However, meeting SAR conditions can be challenging. In this dissertation, a thorough examination of the parameters that affect RF-induced AIMD heating is performed. Novel techniques for mitigating device heating are addressed including a workflow that provides guidance for modifying protocols to reduce the applied RF power with limited impact on diagnostic image quality. Additional work provides an analysis of the magnitude of the RF electric fields with the aim of reducing the induced device heating.

The motivation behind this thesis is provided in Chapter 2. A review of the risks associated with MRI exams for patients with MR-conditional and non-conditional cardiac AIMDs. The current clinical practice for performing MRI examinations is provided in Chapter 3. Chapter 4 is a brief introduction to the physics behind the static magnetic and time varying electromagnetic fields used in MRI. A brief introduction to nuclear magnetic resonance (NMR) is also provided.

Chapter 5 main goal is to find a method to allow MRI clinics to modify standard protocols while limiting the effect on image quality and scan time. To do so, a workflow

iii

that modifies SAR dependent sequence parameters in a specific order was suggested and tested for head, C-spine, L-spine and T-spine vendor provided protocols. The protocols were modified to meet SAR levels as low as 0.1 W/kg which is ~95% less than SAR for conventional examinations. We believe that the proposed workflow will solve the potential examination denials of MR-conditional devices due to the lack of guidance to meet really low SAR conditions.

Chapter 6 presents a simple method for reducing RF induced lead-tip heating (LTH) for patients with MRI unsafe cardiac AIMDs. LTH is known to depend on the lead-path and distance to isocenter. Herein we suggest that by simply changing patient orientation from a head-first to a feet-first orientation LTH can be mitigated. The principle behind this is steeped in the fact that the RF electric field is spatially asymmetric and temperature increase is proportional to the magnitude of the electric field. Thus, by moving to a lower E-field magnitude, LTH can be reduced. Hence, for MR-unsafe cardiac AIMDs implanted on a left side switching from a head-first to a feet-first orientation was then suggested. To do so temperature data was acquired for MR-unsafe cardiac devices following different lead paths and at nine positions within the human body. The results show that changing orientation can reduce LTH. There was no scenario in which LTH was significantly worse in a feet-first orientation.

In Chapter 7 a comparison of LTH at 1.5T and 3T is provided. With 1.5T being the standard field strength used for scanning patients with cardiac AIMDs, examinations at 3T are generally contraindicated. The principal motivation is that SAR increases with the square of field strength, thus the rationale suggests that for patients with cardiac AIMDs, RF-induced device heating will increase as well. Notably, however, the interactions

iv

between the cardiac AIMD and the RF-electromagnetic field are subject to resonant length effects, hence LTH may not be worse. In addition to resonant effects, LTH depends