I am studying automatic segmentation of pathologies in the brain, for which I would need a manually segmented and labelled set of CT brain scans. Does anybody know of such a segmented dataset?
The Cancer Imaging Archive might be worth a try.
http://www.cancerimagingarchive.net/
try downloading osirix, a free imaging platform. lots of good tools there
It is very difficult to see soft tissue structures in CT. An MRI atlas can be found at: http://www.amc.edu/Academic/software/
The Harvard SPL Brain Atlas (MRI) might be useful, since it consists of image files and data sets that are usable by software, rather than being an "application" with a user interface. See "http://www.spl.harvard.edu/publications/item/view/2037". I have played around with converting these from a Slicer-specific to standard DICOM encoding, as DICOM segmentation objects, so if you need me to explain how to do that, let me know ([email protected]). Obviously for an MRI based atlas to be relevant to CT you would need to perform some kind of registration (but than you need to do that from a patient to an atlas anyway, regardless of modality).
- Badges
- Science topic