First of all, I want to express my appreciation to your contribution of FSL and
SIENA algorithm. I am a graduate student doing some brain atrophy
measurement work recently, and got some confusion after reading your paper.
I have questions about the change quantification part. After finding all the
brain edge points in the baseline and repeat images, you resampled them and
stored them in two arrays. here the two arrays have same number elements,
right?
Then you precess the two arrays in two ways, and correlated them to find
the matched points of every points in array2 from array 1.
I got confused about the edge point motions. You maybe calculate the
distance between the two matched corresponding edge points in array 1 and
array 2. As it is stated in your paper :"The position of optimal displacement is
estimated to sub-voxel accuracy by fitting a quadratic through the correlation
values at the peak and its two neighbours". I can't understand the process of
edge point motions calculation, and hope you can give me some hints.
Maybe I have some misunderstandings, and sorry for my erros as a new
learner.
Thank you very much.
Sincerely,
Chunfang
|