Chronological Thread 

< Chronological >      < Thread >

---

Hi Prateep,

Thanks, that helped a lot! I do still have one question on the header info -
how does one determine the offset and why does it matter? If the images I'm
feeding Run are registered already in equal volumes, can I confidently just
set it to zero?

It looks like I'm able to read in mha images that I've pre-processed and Run
begins to distribute points on the surface, but it's consistently failing
silently after only two particle splits. I've attached the full output from
the software below, do you see anything in my parameters that might cause the
particle distribution to fail? Alternately, is there a good method for
generating surface points to feed in (such as randomly selecting
number-of-particles surface points from my binary blob)? I should mention
that the shapes I'm trying to compare are very well registered mid-diaphaseal
scans of mouse tibias. A picture of what one looks like is attached, it's a
pretty smooth surface with several divots and open spaces in it is the idea.
There aren't any encapsulated empty spaces, however, as the MATLAB function
imclose(bw,'holes') was performed on the 3D point cloud before writing.

m_processing_mode = 3
m_number_of_particles = 1024
m_optimization_iterations = 200
m_output_points_prefix = myRun
m_output_transform_file = output_transform_file
m_domains_per_shape = 1
m_timepts_per_subject = 1
m_starting_regularization = 10
m_ending_regularization = 0.1
m_iterations_per_split = 200
m_relative_weighting = 1
m_norm_penalty_weighting = 0
m_initial_relative_weighting = 0.05
m_initial_norm_penalty_weighting = 0
m_adaptivity_strength = 2
m_attributes_per_domain = 0
m_checkpointing_interval = 20
m_transform_file =
m_prefix_transform_file =
m_procrustes_interval = 0
m_recompute_regularization_interval = 1
m_procrustes_scaling = 1
m_adaptivity_mode = 0
m_keep_checkpoints = 0
m_optimization_iterations_completed = 0
Procrustes scaling is on
Mean curvature magnitude = 0.0136963
Std deviation = 0.0109229
Mean curvature magnitude = 0.0139275
Std deviation = 0.00958931

Particle count: 2
........................................................................................................................................................................................................Writing
myRun.0.wpts with 2points
Writing myRun.1.wpts with 2points

Particle count: 4
........................................................................................................................................................................................................Writing
myRun.0.wpts with 4points
Writing myRun.1.wpts with 4points

Particle count: 8
..................................................

Thanks!

+++++++++++++++++++++++++++++
Daniel Leib
Department of Orthopaedic Surgery
Musculoskeletal Research Center
Washington University in St. Louis
Office 11506 BJCIH
314-747-2769 Phone
314-362-0334 Fax


-----Original Message-----
From: prateepm@sci.utah.edu [mailto:prateepm@sci.utah.edu]
Sent: Thursday, September 11, 2014 11:58 AM
To: shapeworks-users@sci.utah.edu; Leib, Daniel
Subject: Re: [shapeworks-users] Getting data into appropriate formats

Hello Dan,

Shapeworks can process volumetric data in mha/nrrdformats. For the NRRD
format, the headers are compliant with Teem
format(http://teem.sourceforge.net/nrrd/index.html). You can find scripts to
save images in this format in MatlabCentral. For the Mha formats, however, no
specific header format is required. Below is the header of a typical mha
volume we use. Please check if your data matches this format.

ObjectType = Image
NDims = 3
BinaryData = True
BinaryDataByteOrderMSB = False
CompressedData = False
TransformMatrix = 1 0 0 0 1 0 0 0 1
Offset = -58 -190 -319
CenterOfRotation = 0 0 0
AnatomicalOrientation = RAI
ElementSpacing = 1 1 1
DimSize = 103 72 121
ElementSize = 1 1 1
ElementType = MET_INT
ElementDataFile = LOCAL


Hope that helps.

Prateep

On Thu, 11 Sep 2014 16:02:55 +0000
"Leib, Daniel" <leibd@wudosis.wustl.edu> wrote:
> Good morning!
>
> I'm hoping to try using ShapeWorks as a means to investigate skull
>shapes in several different mouse phenotypes, and I can run the
>tutorials for ShapeWorks just fine, but I'm having trouble getting from
>my CT scans to an appropriate format.
>
> What I would like to be able to do is go from 16 bit grayscale images
>generated from Scanco's software, preprocessed in MATLAB, and then
>written out to the necessary format for ShapeWorksRun, but I can't seem
>to get any mha or DICOM files with an mhd header to play nice. Are
>there any examples/documentation on what ShapeWroksRun is explicitly
>looking for or what header information must be there to run correctly?
>I've been looking at the ITK documentation and there are so many ways
>to do things that I'm worried I'm just not doing things in the expected
>way.
>
> Thanks,
>
> Dan
>
> +++++++++++++++++++++++++++++
> Daniel Leib
> Department of Orthopaedic Surgery
> Musculoskeletal Research Center
> Washington University in St. Louis
> Office 11506 BJCIH
> 314-747-2769 Phone
> 314-362-0334 Fax
>
>
> ________________________________
>
> The materials in this email are private and may contain Protected
>Health Information. If you are not the intended recipient be advised
>that any unauthorized use, disclosure, copying, distribution or the
>taking of any action in reliance on the contents of this information is
>strictly prohibited. If you have received this email in error, please
>immediately notify the sender via return email.

--
Best
Prateep

________________________________

The materials in this email are private and may contain Protected Health
Information. If you are not the intended recipient be advised that any
unauthorized use, disclosure, copying, distribution or the taking of any
action in reliance on the contents of this information is strictly
prohibited. If you have received this email in error, please immediately
notify the sender via return email.

Attachment: diaphysis.png
Description: diaphysis.png

---

• [shapeworks-users] Getting data into appropriate formats, Leib, Daniel, 09/11/2014
  • Re: [shapeworks-users] Getting data into appropriate formats, prateepm, 09/11/2014
    • RE: [shapeworks-users] Getting data into appropriate formats, Leib, Daniel, 09/11/2014

---

Archive powered by MHonArc 2.6.18.