Section: New Results
Arterial Spin Labelling
Denoising arterial spin labeling MRI using tissue partial volume
Participants : Jan Petr, Jean-Christophe Ferré, Jean-Yves Gauvrit, Christian Barillot.
Arterial spin labeling (ASL) is a noninvasive MRI method that uses magnetically labeled blood to measure cerebral perfusion. Spatial resolution of ASL is relatively small and as a consequence perfusion from different tissue types is mixed in each pixel. An average ratio of gray matter (GM) to white matter (WM) blood flow is 3.2 to 1. Disregarding the partial volume effects (PVE) can thus cause serious errors of perfusion quantification. PVE also complicates spatial filtering of ASL images as apart from noise there is a spatial signal variation due to tissue partial volume. Recently, an algorithm for correcting PVE has been proposed. It represents the measured magnetization as a sum of different tissue magnetizations weighted by their fractional volume in a pixel. With the knowledge of the partial volume obtained from a high-resolution MRI image, it is possible to separate the individual tissue contributions by linear regression on a neighborhood of each pixel. We have proposed an extension of this algorithm by minimizing the total-variation of the tissue specific magnetization. This makes the algorithm more flexible to local changes in perfusion. We show that this method can be used to denoise ASL images without mixing the WM and GM signal.
Improving arterial spin labeling data by temporal filtering
Participants : Jan Petr, Jean-Christophe Ferré, Jean-Yves Gauvrit, Christian Barillot.
Arterial spin labeling (ASL) is an MRI method for imaging brain perfusion by magnetically labeling blood in brain feeding arteries. The perfusion is obtained from the difference between images with and without prior labeling. Image noise is one of the main problems of ASL as the difference is around 0.5-2% of the image magnitude. Usually, 20-40 pairs of images need to be acquired and averaged to reach a satisfactory quality. The images are acquired shortly after the labeling to allow the labeled blood to reach the imaged slice. A sequence of images with multiple delays is more suitable for quantification of the cerebral blood flow as it gives more information about the blood arrival and relaxation. Although the quantification methods are sensitive to noise, no filtering or only Gaussian filtering is used to denoise the data in the temporal domain prior to quantification. We have proposed an efficient way to use the redundancy of information in the time sequence of each pixel to suppress noise. For this purpose, the vectorial NL-means method is adapted to work in the temporal domain. The proposed method is tested on simulated and real 3T MRI data. We have demonstrated a clear improvement of the image quality as well as a better performance compared to Gaussian and normal spatial NL-means filtering.
