Text archives Help

---

Hi Charles,

somehow the human visual system is so heavily dependent on 3D perception that it is almost frustrating to view 2D ortho-slices of anatomy. With no additional information, we cannot resolve our perspective (bottom-up or top-down). The problems are compounded when 3rd party software starts to abuse file format definitions.

All scanner makers have their own conventions. The DICOM standard is fairly reliable. If you can get dicom data directly from the scanner, you should be safe with that data.

In addition, I highly recommend that any MRI scans use some technique to differentiate left and right. I always put a radiological marker on the left cheek bone for all cerebral imaging. If all else fails, I can look for that marker in the data and then I can be sure about the left and right side (so long as the marker was actually put on the left cheek before the scan!).

Without these safety nets, it can be impossible to determine left and right in 2D slices in the anterior-posterior direction (ie, coronal) and the inferior-superior direction (axial or transverse).

I became very annoyed with some software packages that were creating "Analyze" volumes that were not strictly Analyze 7.5 formats. I could not rely on the left-right orientation after my data went into various software. A solution to the problem, for me, was to program my own functions in matlab (I didn't have access to Analyze). My functions have been in the public for a while now and no major problems have been reported. I've had opportunities to verify the functions several times. I don't guarantee they work, but I did take great pains to create them.

I learned that radiological view has +x left, +y anterior, +z superior (LAS). I now imagine the radiological view like this. A doctor is standing at the foot of a patient's bed, viewing an x-ray image on those dark, transparent films. The doctor holds the film up to the light, and peers up at the film to the right and then peers down to the patient in the bed, to imagine this image being the patient. It's best if the film has the patient left on the right side of the film, because the doctor, standing at the foot of the bed, is viewing the patient's left side on the right side of the bed.

Axial is a term that is not necessarily consistently defined. The axis of a magnetic bore is the point that passes down the center of the bore. In that sense, it doesn't define a viewing perspective. In the Analyze data format, the axial view is *radiological*. That view should be differentiated from the *neurological* view. The neurologist often views the head and brain from the top down (imagine the neurosurgeon). So, the radiologists and the neurologist have different views of the patient. Those terms, radiological and neurological, do define a viewing perspective. Axial does not. The axial view can be one or the other. It is also known as the transverse plane. It's a cutting plane, not a view of that plane.

Best, Darren



Charles Murphy wrote:

Darren,
I'm still trying to get Scirun to run on my Apple laptop, but I had a 
troubling discussion with the MRI technichian when I had my knee scanned.She 
maintained that 'axial' meant viewed from 'the top down'. Years ago when I 
did some mentoring work with students, we were working on a Siemens Somatom 
Plus. At that time we were told that 'axial' meant 'from the bottom up'. I 
was able to immediately load my scans into Osirix and based on the 
orientation designations (A-P:ANTERIOR-POSTERIOR;L-R:LEFT-RIGHT;H-F: 
HEAD-FOOT) we could only be viewing from the bottom up. Another cause for 
concern is that doctors offices are routinely sent tiled orthogonal views of 
scans where often orientation can be ambiguous.
Charles Murphy
On Saturday, February 25, 2006, at 03:48AM, Darren Weber 
<dweber@radmail.ucsf.edu> wrote:

Hi Chris,

I have some code and notes about Analyze, see

http://eeg.sourceforge.net/bioelectromagnetism.html#MRI_FEATURES

I don't know anything about Teem i/o functions.

Best, Darren


Chris Butson wrote:

I have noticed a discrepancy in the way Teem handles left/right for MR volumes. I have a 160 slice T1 MR acquired from a Siemens scanner. This particular patient has a very asymmetric brain so left/right are easy to identify from inspection. If I load the patient data into SCIRun either from raw dicom files (using DicomNrrdReader) or from a hdr/img volume (generated using Analyze, loaded into scirun with AnalyzeNrrdReader) then the 3D brain volume is flipped left/right. If I view the same hdr/img volume or dicom series in Analyze then left/right are rendered correctly in 3D.

The radiologists convention dictates that for patients who are head first and supine in the magnet, you should envision looking through the feet so the right side of the brain is on the left side of the screen for axial slices. However, 3D renderings are unambiguous in terms of left/right so no convention is needed. So it is a little odd that the scirun renderings seem to be flipped. More disturbing is the difference in 3D rendering from the same data with different pieces of software. Has anyone else noticed this problem? Is there an explanation for this behavior that I am missing?

Thanks,
Chris

===========================================================================
== The SCIRun Users mailing list: send email to majordomo@sci.utah.edu ==
== for more details. ==
== Please acknowledge use of SCIRun in your papers and reports: ==
== see http://software.sci.utah.edu/scirun-biopse_citation.bib ==
===========================================================================

--

Darren L. Weber, Ph.D.
Visiting Postdoctoral Scholar

Dynamic Neuroimaging Laboratory, Department of Radiology,
University of California, San Francisco,
185 Berry Street, Suite 350, Box 0946,
San Francisco, CA 94107, USA.

Tel: +1 415 353-9444
Fax: +1 415 353-9421
www: http://dnl.ucsf.edu/users/dweber

"To explicate the uses of the brain seems as difficult
a task as to paint the soul, of which it is commonly
said, that it understands all things but itself."
--Thomas Willis (The Anatomy of the Brain and Nerves, 1664)

===========================================================================
== The SCIRun Users mailing list: send email to majordomo@sci.utah.edu   ==
== for more details.                                                     ==
== Please acknowledge use of SCIRun in your papers and reports:          ==
==   see http://software.sci.utah.edu/scirun-biopse_citation.bib         ==
===========================================================================

--

Darren L. Weber, Ph.D.
Visiting Postdoctoral Scholar

Dynamic Neuroimaging Laboratory, Department of Radiology,
University of California, San Francisco,
185 Berry Street, Suite 350, Box 0946,
San Francisco, CA 94107, USA.

Tel: +1 415 353-9444
Fax: +1 415 353-9421
www: http://dnl.ucsf.edu/users/dweber

"To explicate the uses of the brain seems as difficult
a task as to paint the soul, of which it is commonly
said, that it understands all things but itself."
--Thomas Willis (The Anatomy of the Brain and Nerves, 1664)

===========================================================================
== The SCIRun Users mailing list: send email to majordomo@sci.utah.edu   ==
== for more details.                                                     ==
== Please acknowledge use of SCIRun in your papers and reports:          ==
==   see http://software.sci.utah.edu/scirun-biopse_citation.bib         ==
===========================================================================

---

• Re: [SCIRUN-USERS] left/right for MR volumes, Darren Weber, 02/25/2006
  • Re: [SCIRUN-USERS] left/right for MR volumes, Charles Murphy, 02/25/2006
    • Re: [SCIRUN-USERS] left/right for MR volumes, Darren Weber, 02/25/2006
  • Re: [SCIRUN-USERS], Walter Gekelman, 02/25/2006

---

Archive powered by MHonArc 2.6.16.