[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

20031124: IDV for fMRI : file format issue

>From: Bernard Mendiburu <address@hidden>
>Organization: UCLA
>Keywords: 200311182332.hAINWDOb014180 IDV file formats

Bernard-

>The vis5d files are OK in IDV (but the projection bug referenced in previus
>mail)

I'm working on this and will respond tomorrow.

>Unfortunately, the DICOM are flat image files.
>For L levels volumes and T timesteps, you''' have to handle L*T files...
>Our researchers dislike them for this reason

If you know the number of levels and times, this might not be so bad
to create a VisAD object out of.

>The most efficient would be to open Analyse-IMG files.
>It's (one of) the most used file format in medical imagery.
>They're a dual file sheme : 
>One header (.hdr) who describe the data
>One actual data file, flat raw data, slice after slice, timestep after
>timestep.
>
>Does you use the visAD file reading code or did your rewrite it for IDV ?
Could you give me some directions to find the "opening file" code in IVD
>and/or visAD ?

For some things we use the native VisAD readers, for others we use
custome readers.

In VisAD, the visad.data.*.<type>Form classes are the ones that handle
the reading of the files.  For example, visad.data.tiff.TiffForm will
read in a TIFF file, visad.data.biorad.BioRadForm handles the
BioRad .PIC files, etc.

In the IDV, we use the DataSource subclasses in ucar.unidata.data.*
packages to handle the IDV specific needs.  For the Vis5D files,
we use the ucar.unidata.data.grid.Vis5DDataSource class which uses
the visad.data.vis5d.Vis5DFamily class.

>BTW :  
>*I've a kinda bug in the "Add New Data Source" window*
>The file list is not updated at opening.
>I must go up/down in the directory tree to uptade the file list.
>Even after a complete reboot of my machine, just like if the directory
>handle was file-saved somewhere.

Hmm, that is odd.  Please remind me what version of the IDV you are
running on and what operating system.  

Thanks.

Don