NON-CARTESIAN MAGNETIC RESONANCE IMAGING STRATEGIES: IMPROVEMENTS IN ACCELERATED CARDIAC MAGNETIC RESONANCE IMAGING AND TECHNICAL CONSIDERATIONS

magnetic resonance imaging;real-time cardiac magnetic resonance imaging;Biomedical Engineering

Sayin, OzanHerzka, Daniel ;

Johns Hopkins University

关键词: magnetic resonance imaging;    real-time cardiac magnetic resonance imaging;    Biomedical Engineering;   

Others  :  https://jscholarship.library.jhu.edu/bitstream/handle/1774.2/59144/SAYIN-DISSERTATION-2018.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: JOHNS HOPKINS DSpace Repository

【 摘 要 】

Magnetic Resonance Imaging (MRI) is a medical imaging modality that is essential for the imaging of heart because of its unparalleled soft tissue contrast and lack of ionizing radiation. However, the dynamic nature of the heart, together with the motion introduced by respiration, make it a challenging task to generate MRI images without motion-induced artifacts. Traditional approaches deal with physiological motion by acquiring the necessary information to create an image in segments over many heartbeats via electrocardiogram (ECG) gating. Nevertheless, advancements in MRI scanner hardware and image reconstruction techniques, over the past decade, have led to the emergence of real-time MRI acquisitions of cardiac scans where the MRI signal used in reconstructing one image is collected in a single shot. The temporal window during which data is collected from the MRI scanner is extremely short (<50ms) for adequately imaging the heart, and such ;;real-time” accelerated imaging entails the recovery of image information from severely undersampled data.Existing techniques that address the problem of reconstructing images from highly undersampled MRI data come with costs, either in the form of additional MRI scans a priori or aggressive assumptions on the underlying spatiotemporal properties of the object being imaged. In this work, a thorough investigation of one such method that requires a lengthy calibration pre-scan is performed, and novel techniques, which leverages the insights gained from this investigation and incorporates other unprecedented ways of tackling the problem, that facilitate the real-time monitoring of cardiac function, without the inconvenience of a separate calibration scan and assumptions on the statistical properties of the heart’s motion, were developed and evaluated in animal and human subject studies, producing images with comparable quality to existing cardiac MRI techniques. The developed techniques have significant potential of improving the patient’s experience in the clinic, while preserving diagnostic power. They also have the potential to enhance other real-time MRI scenarios such as MRI-guided procedures where a priori calibration scans are infeasible.Advisor and First Reader:Daniel A. Herzka, PhD
Assistant Professor, Biomedical Engineering Johns Hopkins University School of MedicineSecond Reader:Aravindan Kolandaivelu, MD Assistant Professor, Cardiology Johns Hopkins University School of Medicine