<html><head><style type="text/css"><!-- DIV {margin:0px;} --></style></head><body><div style="font-family:times new roman, new york, times, serif;font-size:12pt"><DIV>i have 10 series of dicom image(2D) , i want use those as training set to generate mean shape and mode variation.for a certain 3d stucture</DIV>
<DIV>and then apply Geodesic Shape Prior filter to segment new data</DIV>
<DIV>however, i have some problems with that</DIV>
<DIV>1) i plan to segment each slice, then save as a binary image(png) and convert them to mha file as inputs for PCA Estimator.</DIV>
<DIV> is that correct? but png file will lose dicom information (position, resolution etc), so how to deal with that?</DIV>
<DIV>2) each serie has different resolution and slices, so how to normalize?</DIV>
<DIV> </DIV>
<DIV>if anyone has done this before, pls tell me, thank you very much for any suggestion and help</DIV>
<DIV> </DIV>
<DIV> </DIV></div><br>
</body></html>