fix: Use rigid affine initialization for BOLD mask

[effigies]

This may address #1000. This fix was identified by @lobennett.

The affine initialization can afford to be rigid because the follow-up registration is affine:

niworkflows/niworkflows/data/epi_atlasbased_brainmask.json

Lines 1 to 18 in 000931d

	{
	"winsorize_upper_quantile": 0.98,
	"winsorize_lower_quantile": 0.05,
	"metric": ["Mattes"],
	"metric_weight": [1],
	"radius_or_number_of_bins": [64],
	"transforms": ["Affine"],
	"transform_parameters": [[0.1]],
	"number_of_iterations": [[200]],
	"convergence_window_size": [10],
	"convergence_threshold": [1e-09],
	"sampling_strategy": ["Random", "Random"],
	"smoothing_sigmas": [[2]],
	"sigma_units": ["mm", "mm", "mm"],
	"shrink_factors": [[2]],
	"sampling_percentage": [0.2],
	"use_histogram_matching": [true]
	}

I'm seeing if we can still make a PR against the old LTS branch.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 70.90%. Comparing base (000931d) to head (0ed972d).
⚠️ Report is 10 commits behind head on maint/1.14.x.

Additional details and impacted files

@@               Coverage Diff                @@
##           maint/1.14.x    #1064      +/-   ##
================================================
- Coverage         71.57%   70.90%   -0.68%     
================================================
  Files                89       89              
  Lines              8665     8571      -94     
  Branches           1079      935     -144     
================================================
- Hits               6202     6077     -125     
- Misses             2231     2263      +32     
+ Partials            232      231       -1     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.