Added multi threading

[ArthurRochette]

Added multi-threading to the pydicer/convert/data.py/convert method.
Each patient is now processed in a separate thread, significantly improving conversion speed. By default, the method runs in single-threaded mode to maintain backward compatibility.

[pchlap]

Thanks for this contribution @ArthurRochette! This is something we have been hoping to do for some time (#92).

I notice the pylint step is failing in our actions. Would you be able to check those issues and fix them up? Seems to be some formatting things, I would suggest using black to help fix many of these automatically.

One of the issues I experience when experimenting with multithreading in the past was the log output would become very messy. Have you had a chance to check if the logs output into the .pydicer directly are properly formatted. If they aren't I'm wondering if we need to think of a solution to that before merging this.

Thanks again for the PR!

[ArthurRochette]

Hi Phil,
I've fixed the formatting issue.
While logs and parallelism can be challenging, the Python logging library is thread-safe, ensuring proper output formatting. The logs may appear messy due to interleaved outputs, but this shouldn't impact end users. This trade-off seems inevitable with parallel processing, but I'm open to suggestions for improvement.

[pchlap]

Thanks @ArthurRochette! Sounds good. I will give this a test run over the next few days.

There is just one last thing pylint it complaining about. On line 22 of convert/data.py the get_iterator is still being imported even though it is no longer used. If you could remove that then we can run the tests within this PR.