[Brains-users] Error in loading images

Eugene Zeien eugene-zeien at uiowa.edu
Wed Jan 11 16:10:35 PST 2006


http://www.psychiatry.uiowa.edu/mailman/listinfo

You may wish to browse the brains-users archives.  I don't think there's 
a search function :-(

Cherie Marvel wrote:

>ok, I think I have it now. The trick is to filter -plane coronal. That
>puts scans in the correct orientation and in the correct viewers. This
>process creates scans that look the same whether the header field is set
>to 00 or to 04, so i am using the original 00 setting when loading. The
>only problem is that, as you mentioned, i don't know R from L. I'm going
>to leave it at that for now, not knowing for sure, because it doesn't
>necessarily matter for this particular tracing project.
>
>Thanks a bunch Gene! by the way, is there a user's e-mail list or
>webpage I can access for FAQ's?
>
>have a good one,
>cherie
>							
>
>On Jan 11, 2006 02:12 PM, Eugene Zeien <eugene-zeien at uiowa.edu> wrote:
>
>  
>
>>Cherie,
>>	I do not see any specification for the origin within the Analyze7.5
>>image
>>format document.
>>
>>	4 is "coronal flipped". If this puts coronal->coronal,
>>sagittal->sagittal, etc.
>>I'm quite surprised that 1, "coronal", does not also put things in the
>>correct viewers.
>>Anyway, if you're using the strictAnalyze filter (default reader for
>>.hdr/.img files),
>>you have a few more options available. This will require adding
>>options in the little
>>field "Filter options:" at the bottom of the "Load Image" gui.
>>
>>Valid options are:
>>
>>[-swapAP <swap image anterior/posterior when reading> | off | on]
>>[-swapSI <swap image superior/inferior when reading> | off | on]
>>[-swapRL <swap image right/left when reading> | off | on]
>>[-plane <use specified plane when reading> | none | axial | coronal |
>>      sagittal]
>>
>>combinations are allowed, so you could skip editing the header.
>>
>>-swapSI on -plane coronal
>>-swapSI on -swapAP on -plane coronal
>>-swapRL on -plane axial
>>etc.
>>
>>Caution:  It is very difficult to tell if you have right-left swapped!
>>
>>
>>
>>Cherie Marvel wrote:
>>    
>>
>>>Hi Gene,
>>>
>>>I think that changing the header file helped, but it has not
>>>completely
>>>solved the problem. Using "04" in that specified field yields the
>>>best
>>>results by matching each orientation to the proper viewer (i.e.,
>>>axial
>>>in axial, coronal in coronal, sagittal in sagittal). But--all images
>>>are
>>>180 degrees upside down. At least they are all consistent. So maybe
>>>there is another field in the header that specifies the orientation
>>>of
>>>of the scan in addition to the one I just amended that specifies the
>>>orientation of the plane...???
>>>
>>>
>>>Cherie Marvel, PhD
>>>Postdoctoral Fellow
>>>Johns Hopkins Medical Institutions
>>>Neurology, Div. of Cognitive Neuroscience
>>>1620 McElderry St., Reed Hall East-2
>>>Baltimore, MD 21205
>>>Phone: (410) 502-4664
>>>Fax: (410) 502-2189
>>>www.jhu.edu/nimlab							
>>>
>>>On Jan 11, 2006 10:31 AM, Eugene Zeien <eugene-zeien at uiowa.edu>
>>>wrote:
>>>
>>>
>>>      
>>>
>>>>Cherie,
>>>>
>>>>	Take a close look at the example I included...
>>>>Perhaps I should've reformated this to look more like the text in
>>>>khexedit
>>>>
>>>>0000:00f0 | 00 00 00 00 00 00 00 00 00 00 00 00 01 00 00 00 |
>>>>................
>>>>
>>>>The position that needs to change is where the "1" is above.
>>>>In your file, this line looks like,
>>>>
>>>>0000:00f0 | 20 20 20 20 20 20 20 20 20 20 20 00 00 00 00 00 | .....
>>>>
>>>>To change that field, position the cursor on the preceding 0 and
>>>>type
>>>>01.
>>>>
>>>>0000:00f0 | 20 20 20 20 20 20 20 20 20 20 20 00 [0]0 00 00 00 |
>>>>.....
>>>>0000:00f0 | 20 20 20 20 20 20 20 20 20 20 20 00 0[0] 00 00 00 |
>>>>.....
>>>>0000:00f0 | 20 20 20 20 20 20 20 20 20 20 20 00 01 [0]0 00 00 |
>>>>.....
>>>>
>>>>You may need to try 1 thru 5 in this spot until the orientation
>>>>comes
>>>>out right.
>>>>My guess is that the image is coronal, and 1 will do the trick.
>>>>
>>>>        
>>>>
>>>>>>Try changing the fth line to
>>>>>>
>>>>>>00000f0: 0000 0000 0000 0000 0000 0000 0100 0000 ................
>>>>>>
>>>>>>Save file.  Load image.  Then try,
>>>>>>00000f0: 0000 0000 0000 0000 0000 0000 0200 0000 ................
>>>>>>00000f0: 0000 0000 0000 0000 0000 0000 0300 0000 ................
>>>>>>00000f0: 0000 0000 0000 0000 0000 0000 0400 0000 ................
>>>>>>00000f0: 0000 0000 0000 0000 0000 0000 0500 0000 ................
>>>>>>            
>>>>>>
>>>>Cherie Marvel wrote:
>>>>
>>>>        
>>>>
>>>>>Hi Gene,
>>>>>
>>>>>I was able to view the .hdr file using the handy khexedit function.
>>>>>I'm
>>>>>not sure I understand which lines should be changed (fth line?), so
>>>>>I
>>>>>have attached a snap of my header. Maybe it will make more sense to
>>>>>you.
>>>>>It looks like I have additional columns on the far left, and I
>>>>>don't
>>>>>know what any of these numbers mean. Can you point out which is row
>>>>>f0
>>>>>and 3/4 across? 	
>>>>>
>>>>>thanks for your help on this! --cherie		
>>>>>
>>>>>On Jan 10, 2006 03:02 PM, Eugene Zeien <eugene-zeien at uiowa.edu>
>>>>>wrote:
>>>>>
>>>>>
>>>>>
>>>>>          
>>>>>
>>>>>>Alright.  I found that RedHat includes a neat little hex editor,
>>>>>>khexedit.  Byte 40+108+104=252 should be, row f0 and 3/4 across.
>>>>>>For instance, you'll see something like this:
>>>>>>
>>>>>>0000000: 5c01 0000 4348 4152 0000 0000 0000 0000  ...CHAR........
>>>>>>0000010: 0000 0000 0000 0000 0000 0000 0000 0000  ................
>>>>>>0000020: 0040 0000 0000 7200 0400 8001 8001 e001  . at ....r.........
>>>>>>0000030: 0100 0000 0000 0000 6d6d 0000 2000 0000  ........mm.. ...
>>>>>>0000040: 0000 0000 0001 0200 0800 0000 0000 0000  ................
>>>>>>0000050: 0000 003f 0000 003f 0000 003f 0000 803f  ...?...?...?...?
>>>>>>0000060: 0000 0000 0000 0000 0000 0000 0000 0000  ................
>>>>>>0000070: 0000 803f 0000 0000 0000 0000 0000 0000  ...?............
>>>>>>
>>>>>>Try changing the fth line to
>>>>>>
>>>>>>00000f0: 0000 0000 0000 0000 0000 0000 0100 0000  ................
>>>>>>
>>>>>>Save file.  Load image.  Then try,
>>>>>>00000f0: 0000 0000 0000 0000 0000 0000 0200 0000  ................
>>>>>>00000f0: 0000 0000 0000 0000 0000 0000 0300 0000  ................
>>>>>>00000f0: 0000 0000 0000 0000 0000 0000 0400 0000  ................
>>>>>>00000f0: 0000 0000 0000 0000 0000 0000 0500 0000  ................
>>>>>>
>>>>>>
>>>>>>Cherie Marvel wrote:
>>>>>>
>>>>>>
>>>>>>            
>>>>>>
>>>>>>>Hi Gene,
>>>>>>>
>>>>>>>thanks for looking into the matter. I tried the first suggestion
>>>>>>>of
>>>>>>>applying the old analyze format and got the same results.
>>>>>>>Specifically,
>>>>>>>the axial view shows coronal (and upside down), and the coronal
>>>>>>>view
>>>>>>>shows axial. The sagittal view is sagittal, but S-I and A-P are
>>>>>>>switched
>>>>>>>so that the nose is S at the top of the viewer, and the top of
>>>>>>>the
>>>>>>>head
>>>>>>>is P at the far right of the viewer. the fix may require getting
>>>>>>>messy....
>>>>>>>							
>>>>>>>
>>>>>>>On Jan 10, 2006 12:45 PM, Eugene Zeien <eugene-zeien at uiowa.edu>
>>>>>>>wrote:
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>              
>>>>>>>
>>>>>>>>I'll post this to the "Users" mailing list if it works...
>>>>>>>>
>>>>>>>>I seem to recall an issue similar to this with the "old" analyze
>>>>>>>>format.
>>>>>>>>First thing I'd try, would be to load the image with "analyze"
>>>>>>>>filter
>>>>>>>>specified. The default image filter Brains2 uses is
>>>>>>>>StrictAnalyze.
>>>>>>>>
>>>>>>>>The second option gets a little messy...
>>>>>>>>
>>>>>>>>According to http://www.mayo.edu/bir/PDF/ANALYZE75.pdf
>>>>>>>>the orientation is specified by byte number 104 of the .hdr
>>>>>>>>file.
>>>>>>>>0 = transverse
>>>>>>>>1 = coronal
>>>>>>>>2 = sagittal
>>>>>>>>3 = transverse flipped
>>>>>>>>4 = coronal flipped
>>>>>>>>5 = sagittal flipped
>>>>>>>>The problem is, the .hdr file is binary, so editing it is
>>>>>>>>challenging.
>>>>>>>>I'll go look for a decent binary editor after the 2:00
>>>>>>>>meeting...
>>>>>>>>
>>>>>>>>
>>>>>>>>Cherie Marvel wrote:
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>>                
>>>>>>>>
>>>>>>>>>Hi Gene,
>>>>>>>>>
>>>>>>>>>I'm wondering whether you can help me with something. I have a
>>>>>>>>>very
>>>>>>>>>simple problem loading some images taken here at Hopkins into
>>>>>>>>>BRAINS2. I
>>>>>>>>>hope it has a simple solution. When I load the .hdr images into
>>>>>>>>>BRAINS2,
>>>>>>>>>they show up in the wrong orientations. For example, it says
>>>>>>>>>coronal,
>>>>>>>>>but it looks sagittal. And they are upside-down or turned 90
>>>>>>>>>degress
>>>>>>>>>in
>>>>>>>>>the wrong direction. These were scans acquired by another
>>>>>>>>>researcher
>>>>>>>>>using 1.5T and SPGR. I've asked him to send me as much info as
>>>>>>>>>he
>>>>>>>>>can
>>>>>>>>>regarding the MR sequence used to acquire the images. But I'm
>>>>>>>>>hoping
>>>>>>>>>I
>>>>>>>>>can fix the problem directly with BRAINS2.
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>I scanned the standard work-up manual and the website for input
>>>>>>>>>on
>>>>>>>>>this
>>>>>>>>>particular problem, but what I have found is that all assumes
>>>>>>>>>images
>>>>>>>>>were loaded correctly. Do you have any advice on how to fix
>>>>>>>>>this,
>>>>>>>>>and is
>>>>>>>>>there a webpage that addresses common problems that I may have
>>>>>>>>>overlooked (I think Hans sent out something a while back
>>>>>>>>>directly
>>>>>>>>>people
>>>>>>>>>to problems with BRAINS). If you want, I can recreate the
>>>>>>>>>problem
>>>>>>>>>and
>>>>>>>>>send you the output, if that gives you more information. Just
>>>>>>>>>let
>>>>>>>>>me
>>>>>>>>>know.
>>>>>>>>>
>>>>>>>>>Thank you for your assistance on this. --cherie
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>Cherie Marvel, PhD
>>>>>>>>>Postdoctoral Fellow
>>>>>>>>>Johns Hopkins Medical Institutions
>>>>>>>>>Neurology, Div. of Cognitive Neuroscience
>>>>>>>>>1620 McElderry St., Reed Hall East-2
>>>>>>>>>Baltimore, MD 21205
>>>>>>>>>Phone: (410) 502-4664
>>>>>>>>>Fax: (410) 502-2189
>>>>>>>>>www.jhu.edu/nimlab							
>>>>>>>>>
>>>>>>>>>On Nov 28, 2005 11:43 AM, Eugene Zeien <eugene-zeien at uiowa.edu>
>>>>>>>>>wrote:
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>                  
>>>>>>>>>
>>>>>>>>>>I hope that one slice isn't the full extent of the scan.
>>>>>>>>>>BRAINS2 is intended primarily for working with volumetric
>>>>>>>>>>data.
>>>>>>>>>>Given an image labeled as slice 76, the image reader filter
>>>>>>>>>>(gdcmIPLfilter)
>>>>>>>>>>attempts to locate the remainder of the image set.
>>>>>>>>>>
>>>>>>>>>>Can you place the full set of images on your webserver?
>>>>>>>>>>
>>>>>>>>>>Simone Brockman wrote:
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>                    
>>>>>>>>>>
>>>>>>>>>>>Hello all,
>>>>>>>>>>>
>>>>>>>>>>>We have recently run tinto a problem loading images into
>>>>>>>>>>>Brains2.
>>>>>>>>>>>I
>>>>>>>>>>>suspect that is isn't a problem with the software, but the
>>>>>>>>>>>way
>>>>>>>>>>>in
>>>>>>>>>>>which the images were acquired from the scanner. The
>>>>>>>>>>>technician
>>>>>>>>>>>in
>>>>>>>>>>>Radiology assures me that the same parameters were used for
>>>>>>>>>>>both
>>>>>>>>>>>images and so is at a loss as to why one reads and the other
>>>>>>>>>>>doesn't. (We have had some difficulties previously getting
>>>>>>>>>>>Brains2
>>>>>>>>>>>loaded onto our system however all seems fine now). I
>>>>>>>>>>>wondered
>>>>>>>>>>>if
>>>>>>>>>>>someone in the lab could load the images and see what happens
>>>>>>>>>>>over
>>>>>>>>>>>there. I have made the images available in the following web
>>>>>>>>>>>location
>>>>>>>>>>>
>>>>>>>>>>>http://130.95.22.231
>>>>>>>>>>>Username: download
>>>>>>>>>>>Password: filedown
>>>>>>>>>>>
>>>>>>>>>>>(the image labeled 81635217 is the problem over here)
>>>>>>>>>>>
>>>>>>>>>>>Many thanks for your time and input 
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>Simone Brockman
>>>>>>>>>>>NeuroImaging Laboratory
>>>>>>>>>>>School of Psychiatry and Clinical Neurosciences
>>>>>>>>>>>University of Western Australia
>>>>>>>>>>>Fremantle Hospital
>>>>>>>>>>>Perth, Western Australia
>>>>>>>>>>>+ 61 89431 3643
>>>>>>>>>>>Email: simbroc at cyllene.uwa.edu.au
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>_______________________________________________
>>>>>>>>>>>Brains-users mailing list
>>>>>>>>>>>Brains-users at www.psychiatry.uiowa.edu
>>>>>>>>>>>http://www.psychiatry.uiowa.edu/mailman/listinfo/brains-users
>>>>>>>>>>>
>>>>>>>>>>>                      
>>>>>>>>>>>
>>>>>>>>>>------------------------------------------------------------------------
>>>>>>>>>>                    
>>>>>>>>>>
>
>  
>




More information about the Brains-users mailing list