Draft MEDIC dynamic distortion correction method

[tsalo]

Closes #36.

This currently runs on some test data, but I haven't evaluated the results.

Changes proposed:

• Add sdcflows.workflows.fit.medic module with MEDIC workflow.
  • I translated vanandrew/warpkit into Nipype workflows.

Still to do:

• Figure out how to make EnforceTemporalConsistency more memory-efficient.
  • warpkit is much faster.
• Compare SDCFlows results to MEDIC implementation in warpkit.
  • The current results look really bad compared to warpkit, but I'm confident I can pin down where they diverge.
• Convert functions used as nodes into Nipype interfaces.
• Add MEDIC option to the field map wrangler.
• Documentation
• Tests

[codecov[bot]]

Codecov Report

Attention: Patch coverage is 19.83871% with 497 lines in your changes are missing coverage. Please review.

Project coverage is 66.28%. Comparing base (73d717a) to head (ed0f46c).

Files	Patch %	Lines
sdcflows/utils/misc.py	7.14%	234 Missing :warning:
sdcflows/interfaces/utils.py	24.51%	117 Missing :warning:
sdcflows/workflows/fit/medic.py	12.62%	90 Missing :warning:
sdcflows/interfaces/fmap.py	55.31%	42 Missing :warning:
sdcflows/utils/phasemanip.py	6.66%	14 Missing :warning:

Additional details and impacted files

```diff @@ Coverage Diff @@ ## master #435 +/- ## =========================================== - Coverage 76.43% 66.28% -10.16% =========================================== Files 32 33 +1 Lines 2835 3455 +620 Branches 376 416 +40 =========================================== + Hits 2167 2290 +123 - Misses 600 1097 +497 Partials 68 68 ``` | [Flag](https://app.codecov.io/gh/nipreps/sdcflows/pull/435/flags?src=pr&el=flags&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=nipreps) | Coverage Δ | | |---|---|---| | [unittests](https://app.codecov.io/gh/nipreps/sdcflows/pull/435/flags?src=pr&el=flag&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=nipreps) | `?` | | Flags with carried forward coverage won't be shown. [Click here](https://docs.codecov.io/docs/carryforward-flags?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=nipreps#carryforward-flags-in-the-pull-request-comment) to find out more.

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[tsalo]

I still need to figure out why the results don't look good, but I was thinking an alternative might be for me to simply integrate warpkit as a dependency and run that directly. Would that be preferable to a translated Nipype workflow?

[mgxd]

If using warpkit significantly improves performance, all the more reason to wrap it rather than reimplement - it'll come with the added benefit of reducing longterm maintenance 😄

One hurdle that would add is a more complex installation - however if/when https://github.com/vanandrew/warpkit/issues/6 is resolved it would be a cinch

[tsalo]

@mgxd I did what you proposed in #438, so I'll close this PR now.