[Mristudio-users] How to do a Tailarach transformation?

Christian Bieck c.bieck at T-Online.de
Wed Apr 1 06:08:00 EDT 2009


Hello,

I made an attempt at 2) and did a rigid AIR transformation, followed by an LDDMM. As the grey scale images have to be in byte format, I transformed both the JHU_MNI_GA_DWI sample image and one of my test subject's DWI file to byte format via the BFW button.
I then went on to do an automated single channel LDDMM, not changing the default parameters, and after a while Landmarker said that the files had been sent.
Now while I typed in my registered email address, I didn't get a confirmation nor a Process Complete email, so I'm wondering how long such a remote LDDMM processing normally takes or if I made some mistake during the whole process.
Are there any vital details during LDDMM that I may have overlooked?
Any input would be greatly appreciated. Thanks for your time.

Sincerely,
Christian Bieck  

> 
> Hi Christian,
> 
> Sorry that the manuals for Landmarker and RoiEditor are still quite behind. 
> This year, we will most likely provide 2-day tutorial in Hawaii ISMRM (Sat
> and Sun). Announcement will be sent next week. Join one of these tutorials
> is the best way at this point. We are working on manuals and hopefully can
> provide "Getting Started" soon.
> 
> As for your question, there are several points I want to make;
> 1) There are three ways to do normalization: manual landmark-based (rigid,
> affine, and LDDMM), automated linear & low-order non-linear (AIR), and
> automated high-order non-linear (LDDMM).
> 2) Usually, automated AIR (rigid or affine) + automated LDDMM is the easiest
> way.
> 3) Talairach brain is VERY different from normal subjects. To accurately
> transform your data to Talairach, AIR is not enough. You need extensive
> landmarking or AIR+LDDMM.
> 4) Manual landmark-based normalization has a great flexibility but it's time
> consuming and require some experience.
> 
> I recommend you to start with #2. If you don't come to ISMRM, please wait
> for the Getting Started or you are welcome to visit my lab.
> 
> Susumu




More information about the Mristudio-users mailing list