Segmentation object discussion at tomorrow's hangout

13 messages Options
Segmentation object discussion at tomorrow's hangout – Hi all, We would like to start working on implementing a new "segmentation" node in the Slicer core (with Csaba and the SlicerRT team...
Hi Andras - It would be great if Andrey could join for this discussion. Typically Tuesdays are bad for him, so maybe we should identify a dif...
Steve, Csaba, Andras and I coordinated earlier today. We will have a separate hangout at noon tomorrow, and also I will try to escape my weekly m...
Hi, I suggest we schedule tomorrow hangout to start at 2.30pm. @Andriy: Would that be helpful ? @Andras: Agenda has been updated Thank...
Yes, this is helpful, thank you. I will do my best to join promptly at 2:30. On Mon, Nov 24, 2014 at 5:35 PM, Jean-Christophe Fillion-Robin <j...
Having the hangout at 2:30 tomorrow will work for me. -Steve On Mon, Nov 24, 2014 at 5:48 PM, Andriy Fedorov <fedorov@...> wrote: ...
Hello, This sounds like a very interesting topic on the future of how segmentation in Slicer will behave. I am curious about the memory foot...
Re: Segmentation object discussion at tomorrow's hangout – Hi Brad, It would be great to hear about your ideas. It is clear that a single representation (such as the current labelmap volume) cannot fulf...
I'd love to be in on this discussion, since the anatomy atlas has perhaps the largest number of segments of any common data set, and we'd like...
Re: Segmentation object discussion at tomorrow's hangout – Thanks for the comments, they are very useful. 1-4 are good summary of the API requirements. They all seem to be feasible. Grouping/color ta...
Re: Segmentation object discussion at tomorrow's hangout – Also we should keep in mind that segmentation labels are only one kind of information that we might want to organize similarly. That is, label m...
Yes, the tricky thing is that a general segmentation object could be vector or scalar under the covers, and you'd like that representation det...
Grouping: Node grouping is a general need in MRML and hierarchy nodes work OK already when there are not too many nodes. In case of segmentation, ...