Restore 8acfab54 for impact-reg-konfai 1.8.6

Presets for impact-reg-konfai 1.8.7: one IMPACT loss, sized from their declared cost

Write the configs in KonfAI 1.8.7's spellings

fix: drop ImpactWriteFeatureMaps, which elastix-IMPACT replaced

fix: the ConvexAdam presets run the settings ConvexAdam runs

feat: nine base presets, one per engine and per recommended pairing

feat: no VRAM plan, every preset leaves its patch free and blends its tiles

Drop the config keys konfai 1.8 no longer reads

Write the registration result as Transform.h5 through the itktransform backend

Drop the mask key from Patch, which konfai never binds

A preset declares its displacement field, and only that

Add per-app icons and declare the registration task

Require impact_reg_konfai 1.7.0 for the OME-Zarr displacement field

Pin itk-impact>=0.1.1 in the ConvexAdam presets for the Windows GIL fix

Drop the placeholder checkpoint; these engines are weightless

Move engine code to impact_reg_konfai.models; keep only config + weights

fix(FireANTs): install fireants without its broken simpleitk pin; optional fused kernels (opt-in, pure-torch by default)

Remove the CBCT_CT_TS preset

Ignore __pycache__ and drop tracked bytecode

Add FireANTs presets (Affine/SyN/IMPACT); optional-mask defaults; elastix skips non-restrictive masks

fix(convexadam): disable only the TensorExpr fuser during registration

revert(convexadam): drop the legacy-JIT workaround, fixed model-side

refactor(convexadam): scope the legacy-JIT-executor workaround to the registration

fix(convexadam): disable JIT profiling executor for TorchScript feature models

fix(convexadam): layers_mask as a 1/0 bitmask string, like elastix

feat(convexadam): models as ModelSpec objects for ref/distance pickers

fix(convexadam): quiet the per-patch Registration bar in captured output

fix(convexadam): re-enable autograd for the itk-impact optimisation

refactor: split elastix engine and type model parameters across presets

Remove stale __pycache__/.pyc from preset bundles

Update reg presets: mode knob + FOV evaluator + dimension-aware patch; models.json fetched from HF (unbundled)

Add itk-impact ConvexAdam presets (coarse, fine, composite)

Enable flip test-time augmentation on every preset

Show resolution level and metric in the registration progress bar

Add the generic uncertainty workflow to every preset

Fix MRSeg IMPACT layer mask to match the model (1 layer)

Restructure IMPACT-Reg into per-preset KonfAI apps

Cleanup repository by removing unnecessary files

Update ParameterMap_MRI_TS.txt

Update ParameterMap_MRI_MRSeg.txt

Update ParameterMap_MRI_HN.txt

Update ParameterMap_CBCT_generic_TS.txt

Update ParameterMap_CBCT_generic_MRSeg.txt

Update ParameterMap_CBCT_HN.txt

Update PresetDatabase.json

Update ParameterMap_MRI_HN.txt

Update ParameterMap_MRI_HN.txt

Update ParameterMap_MRI_TS.txt

Update PresetDatabase.json

Update ParameterMap_MRI_TS.txt