Evaluation of Clinical Feasibility of Quantitative MR Perfusion Imaging and Its Calibration

Tissue perfusion is an important physiological parameter that is useful for diagnosis and treatment management of cerebral diseases such as stroke and brain tumors. Using magnetic resonance imaging (MRI), regional cerebral blood flow (CBF) can be estimated with dynamic susceptibility contrast (DSC) or arterial spin labeling (ASL) techniques. DSC-MRI provides robust perfusion evaluations and has been widely used in clinical practice. However, absolute quantification of CBF is difficult with DSC-MRI due to several technical and biophysical limitations such as the inevitable partial volume averaging in the arterial input function (AIF), the uncertain behaviors of the relationships between the intravascular and extravascular relaxivities and the contrast agent concentration and the noise-sensitive calculations. On the other hand, ASL techniques are capable of imaging absolute CBF without the need of contrast agent injection. ASL-MRI has recently evoked a vast amount of clinical interests due to several recent advances that overcame its limitations such as signal-to-noise ratio (SNR), slice coverage and sensitivity to arterial transit delay. Specific methodological developments included pseudo-continuous ASL (PCASL), velocity-selective (VS) ASL, background suppression and fast image acquisitions using 3D and spiral methods. These techniques, some requires proper calibration/scaling, have been shown to be robust in estimating absolute CBF of gray matter in normal subjects. In addition, ASL itself may serve as a calibration medium to turn DSC-MRI into a more quantitative tool. However, studies on optimizations of these methods for clinical use and for tissues with different physiological properties are still limited. The hypothesis of this proposal is that, with careful optimization and calibration/scaling, ASL or combined ASL/DSC MRI can be applied as a clinical routine for estimating absolute CBF. Three particular conditions will be investigated, which includes brain tumor with leaky vessels, neurovascular diseases with delayed arterial supplies and gray and white matter perfusion of aging. Both ASL and DSC-MRI will be acquired in clinical patients for comparison. There are three specific aims targeting to the three conditions: (1) to investigate the potential advantages of ASL over DSC-MRI in brain tumors by comparing ASL results with leakage-corrected DSC results and those acquired using the dual-echo DSC method; (2) to compare VS-ASL with different cut-off velocities, PCASL and ASL-calibrated DSC-MRI with delay-insensitive deconvolution in diseased tissues with arterial transit delays; and (3) to study the influences of background suppression, acquisition mode and signal modeling on PCASL calibration/scaling for obtaining quantitative gray and white matter perfusion in elderly. Toward the end of this project, a set of image acquisition, calculation and calibration methods will be suggested for each of the three conditions with accurate CBF quantification, reasonable image quality and acceptable acquisition time of the MRI scans.

Project ID:PC10301-1141
External Project ID:NSC101-2314-B182-063-MY3