Drop Python 3.9 support and upgrade kwave

[ebrahimebrahim]

Drop python 3.9 support and unpin kwave. See if any disabled tests can then be re-enabled.

Merging this should be followed by a release and then a PR for OpenwaterHealth/SlicerOpenLIFU#486, which realistically would be worked on in parallel.

[dp-openwater]

What are our limitations with sticking with the current version of python and kwave?

What existing issues get resolved by upgrading to python 3.12 and the latest version of kwave?

[ebrahimebrahim]

What are our limitations with sticking with the current version of python and kwave?

What existing issues get resolved by upgrading to python 3.12 and the latest version of kwave?

We are currently pinned to kwave v0.3.4.

kwave v0.3.5 still supports python 3.9, but we had to avoid that one due to a bug (see #193)

Then kwave v0.3.6 onward dropped python3.9 support, so we have been stuck on v0.3.4.

I don't know if there are specific kwave changes that we need yet, as the ultrasound simulation is really Peter's territory. We could dig into it if you want. To see specific kwave changes, see their release notes here: https://github.com/waltsims/k-wave-python/releases
We'd be looking for all updates since v0.3.4

Another point is supporting Slicer 5.10, for the sake of keeping SlicerOpenLIFU available for the community. As of about a week ago, when people download 3D slicer they will download 5.10. We might actually already support Slicer 5.10, or we might be very close to doing so, but we would just need to take a moment to run tests and verify that.

So bottom line is there's no urgency to do all this yet, but it will almost certainly eventually have to be done (python 3.9 is getting old), and with the Slicer 5.10 release last week it finally can be done so now is as good a time as any.

[dp-openwater]

Please go ahead on this. Carry out the update and see if anything breaks. If something does break we revisit and consider whether it's worth the effort to fix.