[Insight-users] Problems with example ImageRegistration8

Thu Jan 19 16:29:40 EST 2006

Hi Aneta,

Your results are not that different from the
ones in the ITK Software Guide, or the ones
that I can compute today with the Nightly
build of ITK.

Here is the output that I got today using
ImageRegistration8 and the two images:

      * brainweb165a10f17.mha
      * brainweb165a10f17Rot10Tx15.mha

Result =
  versor X      = 7.75913e-06
  versor Y      = 2.67549e-05
  versor Z      = -0.0871658
  Translation X = 2.64226
  Translation Y = -17.3228
  Translation Z = -0.00267556
  Iterations    = 24
  Metric value  = 11.8798
Matrix =
0.984804 0.173668 5.19535e-05
-0.173668 0.984804 -2.01234e-05
-5.46588e-05 1.0795e-05 1

Offset =
[-15.0067, 0.00269828, 0.00107133]

They match well what you reported:

 > Versor =
 > 7.7591251791986563e-006 	2.6754909007263262e-005
 > -0.087165786521246499
 > Translation =
 > 2.6422615106837468		-17.322782602080721
 > -0.0026755639690345589

The data in the ITK Software Guide is:

• Versor = (−6.03744e−05,5.91487e−06,−0.0871932)
• Translation = (2.64659,−17.4637,−0.00232496)   mm.

so, even though they are not exactly the same,
they seem to be pretty close to each other.

The rotation is mostly around the Z axis, by an
angle of 10 degrees.

and the translation is only showing variations
of about 1% between the multiple results.

--

About the output volume not being generated correctly:

Could you describe more precisely what do you see in
the resampled volume that make you think it is computed
incorrectly ?

By looking at the resampled volume that I get here,
the only thing that may look like a problem to you
is the fact that there are gray areas around the image.
This is normal, and is caused by the projection of the
fixed image grid to fall outside of the moving image
grid in some sections.

Please give us more details about your concerns with
the resampled image.

      Thanks

         Luis

PS. For your reference,

Here is the progressive trace of the Observer in ImageRegistration8:

0   1097.82   [-0.00217466, -2.62176e-05, -0.00927426, 2.31118,
-15.8114, 0.0692803]
1   1006.57   [-0.00273721, -0.00266639, -0.0272549, 2.33589, -16.002,
0.0276876]
2   804.327   [-0.00397976, -0.00378281, -0.0452322, 2.39116, -16.1872,
-0.00865632]
3   577.771   [-0.00496447, -0.00426568, -0.0620324, 2.46941, -16.3641,
-0.0469688]
4   338.77   [-0.00411104, -0.00378916, -0.0779474, 2.55084, -16.5426,
-0.0684969]
5   121.925   [-0.00193404, -0.00158122, -0.0875469, 2.60193, -16.7347,
-0.0788712]
6   44.6677   [-0.00122303, 0.000595086, -0.0870132, 2.61741, -16.9338,
-0.0867594]
7   27.2755   [-0.000927862, -0.0015867, -0.0880248, 2.66385, -17.1279,
-0.0746458]
8   20.2994   [0.00217117, 0.00702973, -0.0830687, 2.57563, -17.3008,
-0.0309782]
9   79.3637   [-0.00102259, -0.00128144, -0.0880531, 2.6694, -17.3284,
-0.0259501]
10   17.0613   [0.00155747, 0.00179651, -0.0858115, 2.6268, -17.3087,
-0.0113866]
11   21.2675   [-0.00012264, -5.13974e-05, -0.0872728, 2.64733,
-17.3217, -0.0123725]
12   12.0627   [0.000399404, 0.000273191, -0.0868591, 2.63897, -17.3183,
-0.0038612]
13   12.6066   [-0.000107072, -6.69658e-05, -0.0872446, 2.64343,
-17.3224, -0.00462725]
14   11.9945   [0.000151274, 0.000129646, -0.0870653, 2.64105, -17.321,
-0.0033628]
15   12.0766   [2.79253e-05, 3.6039e-05, -0.0871502, 2.64219, -17.322,
-0.00351965]
16   11.9025   [-8.37758e-05, -4.88943e-05, -0.0872272, 2.64318,
-17.3231, -0.00326257]
17   11.9571   [-1.84534e-05, 7.49867e-07, -0.0871832, 2.64258,
-17.3227, -0.00307507]
18   11.8725   [4.4775e-05, 5.00096e-05, -0.0871403, 2.64198, -17.3224,
-0.00272803]
19   11.9222   [1.34112e-05, 2.55414e-05, -0.0871611, 2.64227, -17.3226,
-0.0027949]
20   11.8817   [-1.5931e-05, -5.04805e-07, -0.0871819, 2.64257,
-17.3228, -0.00284077]
21   11.8685   [-1.52052e-06, 1.32087e-05, -0.0871716, 2.64241,
-17.3227, -0.00279268]
22   11.8639   [7.75913e-06, 2.67549e-05, -0.0871658, 2.64226, -17.3228,
-0.00267556]

============================================================================

Markova Aneta wrote:
> Hi all,
> 
> I am trying the ImageRegistration8 example over the images
> brainweb165a10f17.mha and brainweb165a10f17Rot10Tx15.mha as suggested in
> the ITK guide. The example compiles and runs (debug mode) without any
> problem. I am using VS2003 and ITK 2.4.1 (debug mode).
> 
> I have two questions:
> 1. I get results that are quite different from the one in the ITK guide.
> Is such deviation acceptable or not and what could be the possible
> reason?
> 
> Versor =
> 7.7591251791986563e-006 	2.6754909007263262e-005
> -0.087165786521246499
> Translation = 
> 2.6422615106837468		-17.322782602080721
> -0.0026755639690345589
> Matrix = 
> 0.98480424988861470		0.17366803592531921
> 5.1953489460479446e-005
> -0.17366803509494044		0.98480425119985693
> -2.0123410461564691e-005
> -5.4658810456326085e-005	1.0794959713874754e-005
> 0.99999999844794163
> Offset =
> -15.006735272293414		0.0026982755343443075
> 0.0010713291881927489
> 
> numberOfIterations = 24
> bestValue = 11.879821191901589
> 
> 2. It appears that the output volume is not generated correctly (I have
> included a snapshot of the displayed volume). Could anyone give me a
> hint what could be the problem?
> 
>  
> 
> 
> Thankx,
> Aneta
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users