Segmentation filter

4 replies [Last post]
Mon, 06/15/2009 - 13:32
TM
User offline. Last seen 22 weeks 5 days ago. Offline
Joined: 05/26/2009

Which filter to apply in Mimics?

0
Your rating: None
Top
Wed, 07/01/2009 - 11:22
#1
LarsBonitz
User offline. Last seen 12 weeks 2 days ago. Offline
Joined: 06/30/2009
filter
Prerequisite for a valid segmentation is the use of appropriate filters to display only the information to process the necessary and required. For CT and MRI data are good filters in place. Frequently, however, are filters in the field of CBCT still insufficient. By low radiation, the values for discrimination, contrast and noise are higher. This makes the segmentation process especially in the field of dental applications such as SimPlant more difficult. Here we looking for new ways of solutions.
 
Top
Fri, 06/19/2009 - 09:16
#2
Erik Boelen
User offline. Last seen 1 day 23 hours ago. Offline
Joined: 03/13/2009
Image filter

The type of filter you want to use depends on the type of noise you have. Can you add an image to your post to illustrate?
Also you can use combinations of filters to get to the desired result (although the filters can't perform miracles, they can only help to reduce noise to a certain extend).

Top
Wed, 01/27/2010 - 15:27
#3
spagnolo
User offline. Last seen 6 weeks 2 days ago. Offline
Joined: 01/27/2010
MR data- cartilage

Hi,
I am working on MR data, using the 3D LW tool to recostruct the external profile (cortical bone) of the distal femural and of the proximal tibial epiphyses in the knee articulation, including the cartilage.
My goal is to decrease the time needed to obtain a reconstruction meanwhile obtaining precise profiles.  I can include a few extra pixels in small areas but I can't exclude any pixel that can be cartilage or cortical bone. Which combination of filters do you suggest?
I am working on data similar to the one that you can see in the added images.
 
Thank you
Abramo
 

Top
Mon, 02/01/2010 - 12:43
#4
Erik Boelen
User offline. Last seen 1 day 23 hours ago. Offline
Joined: 03/13/2009
MR cartilage

Dear Abramo,
 
Easiest way to improve segmentation results is to optimize the MR protocol to generate higher quality images. Nevertheless, segmenting knee cartilage is in my opinion always a bit subjective; you might include a certain part, whereas someone else will not.
 
Unfortunately you will need to try filters or combinations of filters yourself, I don't know the ideal combination, as I have never tried all possible combinations. Doesn't the median filter give you some better view?
Do keep in mind that the image filters can have an effect on the accuracy of the final 3D.
Another thing you could try is to use the Interpolated view. In the View menu, select Interpolated images. Some people prefer this view over the pixelated view.
 
I assume the new LiveWire is quite helpful for this kind of segmentations?! Would you agree?
Did you watch the Mimics 13.1 video here on the UC? It shows you how best to use it.

Top