This week I discussed MR perfusion imaging with Dr.Gupta. As I am working on estimating the velocity and flow of blood using arterial spin labeling (ASL) , I want to learn more about the clinical status of such perfusion imaging methods. However, I learned that ASL is not performed to patients routinely for many reasons. First, the blood flow information is not needed for most patients. Second, ASL seems not to be as a mature method as other perfusion methods like T2 DSC or T1 DCE perfusion. I compared ASL images with T1 DCE images, and I found the main difference is T1 DCE imaging has a much higher resolution than ASL images, because in T1 DCE imaging we use gadolinium as contrast agent, while in ASL imaging we do not use contrast agent, instead, we use labeled water molecules to examine the velocity of vessels. Labeled water molecules have free diffusion between vessels and tissues, and they will have energy exchange with unlabeled water molecules, causing the T1 decay. This free diffusion and T1 decay will decrease our spatial resolution. But ASL imaging also have its advantage. Gadolinium may cross blood brain barrier and accumulate in your brain, causing cognitive impairment, while ASL imaging don't use contrast agent, so it's safer.
Since I can't watch clinical perfusion imaging during this immersion term, Dr.Gupta suggested me to work with Junghun in Dr.Wang's lab, who is working on cerebral metabolic rate of oxygen (CMRO2) imaging with stroke registry patients. I went through his work and found a common problem we are facing in CMRO2 imaging and velocity estimation is we need a good estimation of cereal blood volume(CBV) and cereal blood flow(CBF). This require a good modeling of vessel structure in a voxel. Till now the vessel structure we assumed in CMRO2 is very simple, we assume there is only one cylinder vessel in one voxel, and this is not accurate, as we may have thousands of capillaries in a 1mm voxel. I am working on building a better model of the vessel structure now.
Next week I will work with Aaron and watch some surgery and talk with Junghun about his CMRO2 work.
No comments:
Post a Comment