Hey Darren,

Is this text version easier to read?  If not let me know!


On Fri, 11 Jul 2003, Darren Weber wrote:

> yep, already using monospace fonts ;-)
> 
> ----- Original Message ----- 
> From: "Evelina" <[EMAIL PROTECTED]>
> To: <[EMAIL PROTECTED]>
> Sent: Friday, July 11, 2003 3:52 PM
> Subject: Re: coordinates conversion
> 
> 
> >
> > Have you tried setting your font to a monospace type face so as to bypass
> > the irritating pdf file transfer aspect of things?
> >
> > Good luck with your analyses!
> >
> > On Fri, 11 Jul 2003, Darren Weber wrote:
> >
> > >
> > > Hi Tosa,
> > >
> > > is there any chance you can forward your mapping diagram as a pdf with
> correct alignment?  It looks very useful, but hard to read in the email.
> > >
> > > Best, Darren
> > >   ----- Original Message ----- 
> > >   From: Yasunari Tosa
> > >   To: [EMAIL PROTECTED]
> > >   Sent: Friday, July 11, 2003 7:35 AM
> > >   Subject: Re: coordinates conversion
> > >
> > >
> > >   Currently coordinate systems for volume, surface, and functional MRI
> are
> > >   very convolved due to historical reason (i.e. originally the volume is
> made of
> > >   coronal direction with 256^3, 1mm voxel size).  We really need to
> simplify all these
> > >   to just one physical RAS and one voxel coordinate systems.   But that
> is not the
> > >   case at this moment.
> > >
> > >   Here is the relations between them.   In order to get from surface RAS
> > >   to the original volume voxel is slightly involved, but a simple matrix
> multiplication.
> > >   (three calculations of transforms and multiplication of matrices;-) ).
> > >
> > >   LInes may not line up but you know what I mean
> > >   (The definition of 0-7 transforms are given below)
> > >
> > >                     0. stored in volume file
> > >        original volume        -------->     RAS ("scanner RAS",
> c_(r,a,s)!=0 in general)
> > >             |                                |
> > >             | 1. calculated                    | identity
> > >             |                                |
> > >             V       2. calculated             V
> > >      conformed volume         -------->     RAS (to have the same
> c_(r,a,s) as above)
> > >         256^3 with 1mm voxel                 |
> > >             |                                |
> > >             | identity                         | 3. calculated (rotation
> and translation)
> > >             |                                |
> > >             V       4. fixed(standard)        V
> > >      conformed volume         -------->     "RAS" where c_(r,a,s) = 0
> > >         256^3 with 1mm voxel                 |
> > >             |                                |
> > >             | identity                         | identity
> > >             |                                |
> > >             V        (same as 4)             V
> > >        "surface voxel"          -------->   Surface coordinates with
> c_(r,a,s) = 0
> > >
> > >   Functional analysis(fMRI) part uses another coordinate system to map
> from the src volume.
> > >
> > >                      5. fixed("standard")
> > >        original volume        -------->     RAS where c_(r,a,s) = 0
> > >             |                                |
> > >             | 6. calculated                    |  mri2fmri (registration
> will give this)
> > >             |                                |
> > >             V        7. fixed("standard")      V
> > >        overlay volume         -------->     RAS where c_(r,a,s) = 0
> > >
> > >   All these coordinate systems make it a rather difficult task to trace
> to the
> > >   original source volume voxel index from surface or functional index.
> > >   If you can follow the arrows, you can get the necessary transforms
> easily.
> > >
> > >   The transform 2 (CORONAL coordinates) is calculated so that the
> following
> > >   equation holds.  That is, the direction cosine part is fixed, but not
> the
> > >   translation part.  In this way, the conformed volume always in the
> CORONAL
> > >   orientation.
> > >                 [-1  0  0 s1][128]   [c_r]
> > >                 [ 0  0  1 s2][128] = [c_a]
> > >                 [ 0 -1  0 s3][128]   [c_s]
> > >                 [ 0  0  0  1][ 1 ]    [ 1 ]
> > >
> > >   where c_(r,a,s) is from the "scanner RAS".  This "scanner RAS" has the
> physical
> > >   meaning of "Right-Anterior-Superior" directions of the head.
> > >
> > >   The transform 4 is fixed as
> > >                 [-1  0  0  128]
> > >                 [ 0  0  1 -128]
> > >                 [ 0 -1  0  128]
> > >                 [ 0  0  0   1 ]
> > >
> > >   The transform 1 is calculated by (in a matrix sense)
> > >
> > >        xform1 = inv(xform2) * xform0
> > >
> > >   The transform 3 is calculated by (in a matrix sense)
> > >
> > >        xform3 = xform4 * inv(xform2)
> > >
> > >   The transform 5 and the transform 7 are calculated by the requirement
> > >
> > >                 [-1  0  0 s1][xsize  0     0     0][width/2 ]   [0]
> > >                 [ 0  0  1 s2][  0  ysize   0     0][height/2] = [0]
> > >                 [ 0 -1  0 s3][  0    0   zsize   0][depth/2 ]   [0]
> > >                 [ 0  0  0  1][  0    0     0    1][    1   ]   [1]
> > >
> > >   Here the name "RAS" lost the meaning completely.  This "RAS" is just
> to be used
> > >   for alignment purpose only.  The width/height/depth are for the
> appropriate volume.
> > >   The original volume could be sagittal or horizontal, but it uses the
> coronal
> > >   transformation to be the "RAS".
> > >
> > >   Tosa
> > >
> > >   On Fri, 2003-07-11 at 03:57, Denis Fize wrote:
> > > Hello,
> > >
> > > The label utility is a very nice feature of freesurfer.
> > > One of the usage here is to obtain the volumic coordinates of pieces of
> > > cortex.
> > > Once having the ras coordinates, I would like to convert them back to
> > > the coordinate system of the software used for the statistical analysis.
> > >
> > > Does such tool exists / is planned ? a kind a mri_convert for labels.
> > >
> > > Alternatively, is there a straightforward trick to convert these
> > > coordinates ?
> > >
> > > Thanks,
> > >
> > > Denis
> > >
> >
> > -- 
> > Ev
> >
> >
> >
> 

-- 
Ev 

From: Yasunari Tosa
To: [EMAIL PROTECTED]
Sent: Friday, July 11, 2003 7:35 AM
Subject: Re: coordinates conversion

Currently coordinate systems for volume, surface, and functional MRI are
very convolved due to historical reason (i.e. originally the volume is made of
coronal direction with 256^3, 1mm voxel size).� We really need to simplify all these
to just one physical RAS and one voxel coordinate systems.�� But that is not the
case at this moment.��

Here is the relations between them.�� In order to get from surface RAS
to the original volume voxel is slightly involved, but a simple matrix
multiplication.
(three calculations of transforms and multiplication of matrices;-) ).

LInes may not line up but you know what I mean
(The definition of 0-7 transforms are given below)

����������������� 0. stored in volume file����
���� original volume������� -------->���� RAS ("scanner RAS", c_(r,a,s)!=0 in
general)
��������� |������������������������������� |
��������� | 1. calculated������������������| identity
��������� |������������������������������� |
��������� V������ 2. calculated������������V
�� conformed volume�������� -------->���� RAS (to have the same c_(r,a,s) as above)
����� 256^3 with 1mm voxel���������������� |
��������� |������������������������������� |
��������� | identity�����������������������| 3. calculated (rotation and
translation)
��������� |������������������������������� |
��������� V������ 4. fixed(standard)�������V
�� conformed volume�������� -------->���� "RAS" where c_(r,a,s) = 0
����� 256^3 with 1mm voxel���������������� |
��������� |������������������������������� |
��������� | identity�����������������������| identity
��������� |������������������������������� |
��������� V������� (same as 4)������������ V
���� "surface voxel"��������� -------->�� Surface coordinates with c_(r,a,s) = 0

Functional analysis(fMRI) part uses another coordinate system to map from the src
volume.

������������������ 5. fixed("standard")
���� original volume������� -------->���� RAS where c_(r,a,s) = 0
��������� |������������������������������� |
��������� | 6. calculated������������������|� mri2fmri (registration will give
this)
��������� |������������������������������� |
��������� V������� 7. fixed("standard")����V�����������������������
���� overlay volume�������� -------->���� RAS where c_(r,a,s) = 0

All these coordinate systems make it a rather difficult task to trace to the
original source volume voxel index from surface or functional index.
If you can follow the arrows, you can get the necessary transforms easily.

The transform 2 (CORONAL coordinates) is calculated so that the following
equation holds.� That is, the direction cosine part is fixed, but not the
translation part.� In this way, the conformed volume always in the CORONAL
orientation.�
������������� [-1� 0� 0 s1][128]�� [c_r]
orientation.�
������������� [-1� 0� 0 s1][128]�� [c_r]
������������� [ 0� 0� 1 s2][128] = [c_a]
������� ����� [ 0 -1� 0 s3][128]�� [c_s]
������������� [ 0� 0� 0� 1][ 1 ]���[ 1 ]

where c_(r,a,s) is from the "scanner RAS".� This "scanner RAS" has the physical
meaning of "Right-Anterior-Superior" directions of the head.

The transform 4 is fixed as
������������� [-1� 0� 0� 128]
������������� [ 0� 0� 1 -128]
������� ����� [ 0 -1� 0� 128]
������������� [ 0� 0� 0�� 1 ]

The transform 1 is calculated by (in a matrix sense)

���� xform1 = inv(xform2) * xform0

The transform 3 is calculated by (in a matrix sense)

���� xform3 = xform4 * inv(xform2)

The transform 5 and the transform 7 are calculated by the requirement

������������� [-1� 0� 0 s1][xsize� 0���� 0���� 0][width/2 ]�� [0]
������������� [ 0� 0� 1 s2][� 0� ysize�� 0���� 0][height/2] = [0]
������� ����� [ 0 -1� 0 s3][� 0��� 0�� zsize�� 0][depth/2 ]�� [0]
������������� [ 0� 0� 0� 1][� 0��� 0���� 0���  1][��� 1�� ]�� [1]

Here the name "RAS" lost the meaning completely.� This "RAS" is just to be used
for alignment purpose only.� The width/height/depth are for the appropriate volume.
The original volume could be sagittal or horizontal, but it uses the coronal
transformation to be the "RAS".�

Tosa






Reply via email to