I'm PhD student, I'm working on MRI segmentation, (trying to use 536 & 126 data file of IBSR), and I would like to ask if any one can help about how to get the manual segmentation of tumor (the gold standard)of 536 & 126 data file of IBSR? Because, for 536 data file of IBSR, all what I've on the website is:
the image files:
tar xf 536_32_img.tar
tar xf 536_45_img.tar
tar xf 536_47_img.tar
tar xf 536_68_img.tar
tar xf 536_88_img.tar
the outline files:
tar xf 536_32_otl.tar
tar xf 536_45_otl.tar
tar xf 536_47_otl.tar
tar xf 536_68_otl.tar
tar xf 536_88_otl.tar
However the outlines files doesn't contain manual segmentation of the TUMOR that I need but lists of points that define certain structures in each scan image after semi-automated segmentation techniques.
The same problem I've faced while trying to use the 126 data file of IBSR (brain tumor),
you find an explaning example of image attached below: 536_32_20 with its correspondig outline file: sdl20 otl zoomed 2(no tumor manual segmentation included),
Every one, hanks for considering my questions,
Dear Yahia,
ISBR should not be your best choice for brain tumor segmentation. I strongly recommend you to use BRATS dataset. It has many cases with T1, T2, T1c, and FlAIR MRI for each and comes with ground truth. Also includes low grade and high grade tumors.
Good Luck
I'm also working on BRATS dataset it's more efficient rather than IBSR,
Dear Elazab & Iqbal,
I thank you for answering me. I did check BRATS dataset before and I agree that's efficient as a database for tumor segmentation application, however I would like to use more then one database that's why I'm still looking for 356 & 126 IBSR manual segmentation,
Cordialy
- Badges
- Science topic
- Similar topics
- Computer Science
- Images