The heatmap & output from FAST can it be imported to DSA? #15
Replies: 1 comment
-
Hello, @lomshabhishek! Hope you are doing well! This is something I have been meaning to look into myself. In practice I do not believe it should be that difficult to do, but there are some caveats. First of all, the format used in FAST/FP when storing the output varies between models. For pw-classifiers we store the result in HDF5 format. For pw-emantic segmentation we store the result as pyramidal TIFFs. DSA uses a different way of storing annotations, through GeoJSON, if I recall correctly, which is similar to what QuPath does. Thus, to import annotations Into DSA from FP, requires that you convert the annotation first. In the FP-DSA plugin, when you run a FAST model we actually do the full ordinary FAST pipeline and then convert the result. Thus to only import the result from FAST/FP into DSA, we could simply add a method that does the last part of the plugin. But we should try to support both HDF5 and TIFFs from the get-go. Sadly this often tends to take more time actually developing than I anticipate, and I do not have a laptop where FAST in Docker works atm, but I may have later today. Let's keep this Discussion open and I will update you on any progress :] |
Beta Was this translation helpful? Give feedback.
-
I was wondering if someone can import and visualise results from certain project in FAST into DSA
Beta Was this translation helpful? Give feedback.
All reactions