ai969 at Freenet.carleton.ca (Stev Thomas) writes:
> As a neuroscientist gone programmer (towards becoming a
>computatinal neuroscientist) I want to develop a viewer for MRI
> What I would like to know, from as many respondants as
> 1. Where can I get information about the data format of
>the MRI data file.
> 2. Where can I get some sample data
> 3. What is the type/bran of scanner you use
> 4. What sort of software do you use to view
> 5. Is any of this on the public domain
> Is anyone interested in a cooperative development
I have some experiences of programming for MRI data. Maybe the format
depends on the machine which you are using.
I have used Siemens Magnetom(1.5T), and GE Signa Advantage(1.5T). Both
of them have simple data structures; each pixel consists of 2 bytes
expressing pixel value. The order of the pixels is from the left to the
right, and from the upper line to the bottom line. They have headers
for each images. I cannot tell the size of the headers offhand, but
you can easily get it if you express the size of the file in hexadecimal
notation. For eg, the size is 0x24000, the size of the header is
0x4000 and the image data size is 0x20000(probably 256 * 256 * 2bytes).
But this is just an example. The size of the header is constant.
I'm using my original program on X Windows. If you are familiar with
X Windows or other graphic systems, it is not so hard to program. I don't
know about public domain software of this sort. If you have specific
questions, I should be able to help you.
As for PET, we use Shimazu Headtome IV, and its format is also simple.
But I suppose your system should be different.
Hope this helps.
Tetsuji Rai MD trai at twics.co.jp, 74610.1537 at compuserve.com
Department of Radiology, Hospital attached to University of Tokyo
Home: 5-12-21-301, Toyotamakita, Nerimaku, Tokyo 176, Japan