[Nirfast] NIRFASTSlicer Modules

Vincent Kitsmiller Vincent.J.Kitsmiller.1 at nd.edu
Wed Oct 26 11:16:40 EDT 2016


Thanks for the response. I then have a follow up question. I have a
set of tifs from an MRI. It has been segmented into two regions and
saved as two sets of greyscale and RGB tif files (fibroglandular
greyscale and RGB, adipose greyscale and RGB). Is there a way to
create a region segmented mesh from this data? Thanks.

-Vince

On Wed, Oct 26, 2016 at 10:23 AM, Scott C. Davis
<Scott.C.Davis at dartmouth.edu> wrote:
> Hi Vince,
> When that paper was published, the NIRFAST front end (called NIRView) had been built on an older platform (Volview) which has since been deprecated. Over time, it became very difficult to keep the program up to date and functioning on new operating systems, and the platform limited development of new functionality.  In early 2016, we released NIRFASTSlicer to replace the antiquated Volview-based front end.  This version is built on the 3DSlicer platform, a very active open source project whose capabilities continue to grow.  While overall the platform switch has improved stability, functionality and the ability to continue innovative development, some specific functionality available in the pre-Slicer version is not available in the Slicer package. However, developers are working on improved segmentation tools which I expect will be released soon.
>
> To address the specific issue you raise:  Bias correction, skin extraction, and K-means are not available in the NIRFASTSlicer Editor module.  K-means was a semi-auto segmentation process which did a fair job on breast MR images.  Bias correction (basically a high-pass filter for low frequency noise in MR images) and skin extraction were image processing filters applied to the MR images which improved the results of the K-means segmentation.  These are not available in NIRFASTSlicer, however, we have found that the “GrowCut” function applied to breast MRI does a very good job segmenting the tissue without the need for bias correction and skin extraction.  The tutorial should explain how to use this function on breast MR.
>
> I hope this helps explain.
>
> Best,
> Scott
>
>
> On 10/20/16, 12:58 PM, "Nirfast on behalf of Vincent Kitsmiller" <nirfast-bounces at public.kitware.com on behalf of Vincent.J.Kitsmiller.1 at nd.edu> wrote:
>
>     In reviewing, Fast segmentation and high-quality three-dimensional
>     volume mesh creation from medical images for diffuse optical
>     tomography, I noticed multiple references to modules which I cannot
>     seem to locate in the slicer. I specifically would like to know about
>     those from the breast example, seen in the attached picture. If I
>     could get some direction on where to find these modules that would be
>     wonderful. Thank you.
>
>     -Vince Kitsmiller
>
>


More information about the Nirfast mailing list