The main reason for all this is speed,
the scans are delivered or in bmp(24bits/px) or in jpeg files.
Or I can write some software to convert the loaded 24 bits images to 8 bit images in memory.
So, I'm going for bmp files.
Second is the calculation time, if I need to load every time all the 2D scan images,
and convert them to 3D images.
With my approach I need about 2 minutes to convert the 2D images to one 3D image and the filtered images.
(see below)
I need also pixel manipulation to make them transparent and/or filter them away.
The right images are the CT Scan 2D gray-scale images (149),
so the user can select the start image from where to build the 3D images and filter images.
With this example, we start at image 65.
On the left, the 3D created and filtered image.
So What I think is that 32 bit/px data is what I need to filter the data ??

- every professional was once an amateur - greetings from Pajottenland - Belgium -
PS: sorry for my english I speak flemish ...