spinestats: version 3 — endplate-plane Cobb + apex + tightened multi-Cobb by robert-graf · Pull Request #154 · Hendrik-code/TPTBox

robert-graf · 2026-09-22T11:04:59Z

Summary

Extends the endplate-plane direction machinery (used for v2 lordosis/kyphosis) to the coronal Cobb angle, bumps CURRENT_VERSION to 3 so cached stat files recompute the angle keys, and adds per-curve apex vertebra keys.

Stacks on top of these already-(mostly-)merged branches:

dicom-extract-cleanup
nako-follow-up-export
poi-label-names

Changes

TPTBox/spine/spinestats/angles.py

_get_norm routes Vertebra_Direction_Right through a new _endplate_r_direction helper (analogue of _endplate_ap_direction), so Cobb lines are true endplate lines averaged across the shared disc.
plot_cobb_angle always uses the endplate right direction — the old use_ivd_direction branch used Vertebra_Disc_Inferior from each vertebra separately, giving two different lines at one anatomic disc.
compute_angel_between_two_points_ orders id1/id2 by Vertebra_Instance.order_dict() rather than by numeric label, so T13 (label 28, anatomically between T12 and L1) pairs to the right endplates.
compute_max_cobb_angle_multi uses an exclusive split (+ 1 on the below slice) so sibling curves cannot share an endpoint vertebra.
VERT_START_COBB is C3 (C2 excluded from Cobb search).
compute_lordosis_and_kyphosis emits {curve}_apex keys via a new _find_curve_apex helper; plot_compute_lordosis_and_kyphosis renders a *apex marker.
plot_cobb_angle anchors the label at the apex disc and widens the coronal pad from 50 to 80 mm.

TPTBox/spine/spinestats/_run_all.py

New POI_INVALIDATING_VERSION = 2 constant replaces a bare literal 2 in _poi_is_stale_wrt_stat; its docstring now references a symbol that actually exists (previously pointed at a nonexistent ANGLES_VERSION).
Documents the v3 bump in the CURRENT_VERSION history comment.

TPTBox/spine/spinestats/veridah_angles.py

_compute_k_variant also returns lumbar_lordosis_apex so the per-k variants carry the new apex key.

Test plan

pytest unit_tests/
Run _run_all.py in test=True mode on a NAKO subject and confirm cobb/curv keys, {curve}_apex values, and that a v2 stat file is recomputed to v3 without invalidating POIs.
Eyeball a plot_cobb_angle snapshot and a plot_compute_lordosis_and_kyphosis snapshot to verify the apex marker and label placement.

🤖 Generated with Claude Code

…-Reg

Roll the internal-scripts changes together with the new precompute_grid_info_parallel helper in _load_nako_wh, which fans per-file _add_grid_info_to_json calls out to a ProcessPoolExecutor so the sidecar grid cache is populated up front.

…-Reg

Adds a small registration mechanism so producers can attach per-vertebra direction vectors (endplate PCA normals, orientation vectors, ...) or label-keyed scalars (endplate_internal_angle, curvature_*) to poi.info and have them stay aligned with the POI points automatically: - New TPTBox/core/poi_fun/vector_fields.py with the constants POI_INFO_VECTOR_FIELDS_KEY / POI_INFO_LABEL_KEYED_FIELDS_KEY and three helpers (_transform_direction_vectors_inplace, _rotate_direction_vectors_inplace, _remap_vector_field_keys_inplace). - POI.reorient signed-permutes vectors; POI.resample_from_to applies the real R_ref.T @ R_self rotation; POI.rescale is a no-op. - POI_Global.to_cord_system flips vector components on RAS<->LPS. - POI.map_labels remaps top-level keys of both registries. - endplates.py registers angle_superior/inferior_endplate, endplate_internal_angle, and curvature_* when they're produced. - Fix: vertebra_pois_non_centroids.py now (re)runs calc_endplate_points_ when S1 is present but its Sacrum_Endplate- derived superior-endplate landmark is missing, so the sacrum endplate direction actually lands in the buffer. - Angles: _endplate_ap_direction preferred over Vertebra_Direction_ Posterior for MoveTo.TOP/BOTTOM in _get_norm, yielding endplate- aligned Cobb-style lordosis/kyphosis lines. Docs: new section in docs/api/poi_fun.md. Tests: unit_tests/test_poi_vector_fields.py (13 tests). Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

Merges the poi-label-names branch (LABEL_NAME + nested {region: {subregion: name, "name": group}} format in info["label_name"], new label_name / set_label_name / level_one_name / set_level_one_name accessors, migration helpers, and level_*_info propagation in copy / _set_inplace / calc_poi_average / POI_Global.to_other). Then wires label_name into the auto-transform mapping scheme: - New vector_fields._remap_label_name_inplace handles the nested format: remaps top-level region keys via label_map_region and inner subregion keys via label_map_subregion, preserving the "name" group entry. Colliding regions merge with last-write-wins on inner keys. - map_labels calls both _remap_vector_field_keys_inplace (for the registered flat vector / label-keyed fields) and the new _remap_label_name_inplace on all three return paths. - Also triggers the legacy flat-format migration on remap. Docs updated with a "Special-cased field" section for label_name. Tests: 5 new cases in Test_LabelName_MapLabels covering region-only, subregion-only, combined, region collision, and legacy-format migration paths (18/18 passing). Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

…Cobb Extends the endplate-plane direction machinery used for v2 lordosis/kyphosis to the coronal Cobb angle, and bumps ``CURRENT_VERSION`` to 3 so cached stat files recompute the angle keys (POI buffers stay valid — no new landmarks). Angles (``TPTBox/spine/spinestats/angles.py``): - ``_get_norm`` now routes ``Vertebra_Direction_Right`` through a new ``_endplate_r_direction`` helper (analogue of ``_endplate_ap_direction``), so Cobb lines are true endplate lines averaged across the shared disc instead of the two flanking vertebrae disagreeing on the same boundary. - ``plot_cobb_angle`` always uses that endplate right direction — the ``use_ivd_direction`` branch used ``Vertebra_Disc_Inferior`` from each vertebra separately, giving two different lines at one anatomic disc. - ``compute_angel_between_two_points_`` orders ``id1``/``id2`` by ``Vertebra_Instance.order_dict()`` rather than by numeric label, so T13 (label 28, anatomically between T12 and L1) pairs to the right endplates and ``MoveTo.TOP``/``BOTTOM`` are not swapped. - ``compute_max_cobb_angle_multi`` uses an exclusive split (``+ 1`` on the ``below`` slice) so sibling curves cannot share an endpoint vertebra. - ``VERT_START_COBB`` is C3 (C2 excluded from Cobb search). - ``compute_lordosis_and_kyphosis`` emits ``{curve}_apex`` keys, computed by the new ``_find_curve_apex`` helper (same bisector heuristic as ``compute_max_cobb_angle``). ``plot_compute_lordosis_and_kyphosis`` renders a ``*apex`` marker at that vertebra. - ``plot_cobb_angle`` anchors the label at the apex disc and widens the coronal pad from 50 to 80 mm so labels don't get clipped. Runner (``TPTBox/spine/spinestats/_run_all.py``): - Introduces ``POI_INVALIDATING_VERSION = 2`` so the ``_poi_is_stale_wrt_stat`` threshold is a named constant rather than a magic ``2``, and its docstring points at a symbol that actually exists (previously referenced a nonexistent ``ANGLES_VERSION``). - Documents the v3 bump in the ``CURRENT_VERSION`` history comment. Veridah (``TPTBox/spine/spinestats/veridah_angles.py``): - ``_compute_k_variant`` also returns ``lumbar_lordosis_apex`` so the per-k variants carry the new apex key. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

github-actions · 2026-09-22T11:07:56Z

Benchmark comparison

Speed (wall time per call)

baseline 7d95bc9 vs head 7d95bc9 · python 3.11.17 · 5 repeats + 1 warmup

Showing the 5 most-changed measurements per case; the rest are collapsed. Rows with a baseline below 3 ms are tagged (noise) — runner jitter on those swamps any real change.

`ct_3d` — shape (73, 47, 73)

Measurement	baseline ms (median ±½·range)	head ms (median ±½·range)	Δ %	p
`poi_calc_centroids_nocrop`	4.61 ±0.26	4.21 ±0.10	-8.7%	0.042
`nii_rescale_seg`	3.41 ±0.18	3.19 ±0.08	-6.6%	0.037
`nii_fill_holes`	4.46 ±0.07	4.58 ±0.06	+2.8%	0.083
`poi_calc_poi_from_subreg_vert`	12.91 ±0.78	13.18 ±0.63	+2.1%	0.910
`nii_filter_connected_components`	4.42 ±0.12	4.51 ±0.15	+1.9%	0.115

… 31 more measurements

Measurement	baseline ms (median ±½·range)	head ms (median ±½·range)	Δ %	p
`nii_erode_msk`	6.21 ±0.03	6.32 ±0.03	+1.8%	0.000
`nii_save`	11.85 ±0.16	12.02 ±0.13	+1.5%	0.069
`nii_resample_from_to`	42.06 ±0.22	42.60 ±0.19	+1.3%	0.002
`nii_rescale`	39.89 ±0.20	40.36 ±0.29	+1.2%	0.013
`nii_dilate_msk`	126.19 ±0.45	126.70 ±0.62	+0.4%	0.071
`nii_load_img`	10.93 ±0.77	10.93 ±0.06	+0.1%	0.390
`metric_voxels`	250463.00 ±0.00	250463.00 ±0.00	+0.0%	—
`metric_labels`	3.00 ±0.00	3.00 ±0.00	+0.0%	—
`metric_foreground_pct`	14.51 ±0.00	14.51 ±0.00	+0.0%	—
`poi_save`	0.53 ±0.02	0.35 ±0.01	-34.6% (noise)	0.000
`poi_to_global`	0.21 ±0.03	0.14 ±0.01	-30.2% (noise)	0.005
`poi_resample_from_to`	0.77 ±0.04	0.59 ±0.02	-23.9% (noise)	0.000
`poi_map_labels`	0.20 ±0.01	0.15 ±0.01	-22.7% (noise)	0.000
`poi_load`	1.09 ±0.04	0.89 ±0.05	-18.2% (noise)	0.001
`poi_local_to_global_arr`	0.32 ±0.05	0.27 ±0.01	-15.9% (noise)	0.042
`poi_rescale`	0.20 ±0.02	0.17 ±0.01	-15.7% (noise)	0.051
`nii_load_seg`	2.75 ±0.42	3.09 ±0.12	+12.3% (noise)	0.816
`poi_reorient`	0.37 ±0.03	0.33 ±0.02	-10.6% (noise)	0.026
`nii_apply_crop`	0.69 ±0.04	0.74 ±0.02	+6.9% (noise)	0.015
`nii_reorient`	0.90 ±0.07	0.85 ±0.08	-6.0% (noise)	0.954
`nii_extract_label`	0.43 ±0.02	0.45 ±0.05	+5.0% (noise)	0.885
`nii_get_connected_components`	2.28 ±0.27	2.37 ±0.09	+3.8% (noise)	0.570
`nii_get_array`	0.05 ±0.01	0.05 ±0.00	-3.8% (noise)	0.334
`nii_set_dtype`	0.53 ±0.05	0.55 ±0.05	+3.5% (noise)	0.355
`poi_calc_centroids`	2.43 ±0.08	2.37 ±0.02	-2.2% (noise)	0.088
`nii_volumes`	1.87 ±0.01	1.90 ±0.01	+2.0% (noise)	0.012
`nii_unique`	0.69 ±0.03	0.70 ±0.01	+1.6% (noise)	0.871
`nii_compute_crop`	0.33 ±0.01	0.33 ±0.02	+0.9% (noise)	0.286
`nii_center_of_masses`	1.75 ±0.03	1.76 ±0.01	+0.8% (noise)	0.584
`nii_pad_to`	0.70 ±0.05	0.70 ±0.06	-0.5% (noise)	0.258
`nii_map_labels`	1.09 ±0.03	1.09 ±0.03	-0.1% (noise)	0.460

`ct_2d` — shape (73, 47, 1)

Measurement	baseline ms (median ±½·range)	head ms (median ±½·range)	Δ %	p
`nii_resample_from_to`	7.61 ±0.20	8.01 ±0.74	+5.3%	0.111
`nii_rescale`	5.64 ±0.10	5.71 ±0.18	+1.2%	0.868
`nii_dilate_msk`	14.89 ±0.08	14.89 ±0.08	+0.0%	0.603
`metric_voxels`	3431.00 ±0.00	3431.00 ±0.00	+0.0%	—
`metric_labels`	3.00 ±0.00	3.00 ±0.00	+0.0%	—

… 30 more measurements

Measurement	baseline ms (median ±½·range)	head ms (median ±½·range)	Δ %	p
`metric_foreground_pct`	37.60 ±0.00	37.60 ±0.00	+0.0%	—
`nii_unique`	0.15 ±0.06	0.13 ±0.00	-14.6% (noise)	0.100
`poi_save`	0.55 ±0.08	0.47 ±0.02	-14.0% (noise)	0.110
`nii_get_array`	0.02 ±0.00	0.02 ±0.00	+13.0% (noise)	0.170
`nii_center_of_masses`	0.25 ±0.02	0.22 ±0.01	-11.9% (noise)	0.036
`nii_apply_crop`	0.69 ±0.03	0.75 ±0.04	+8.1% (noise)	0.190
`poi_map_labels`	0.22 ±0.01	0.20 ±0.01	-7.9% (noise)	0.003
`nii_get_connected_components`	0.70 ±0.02	0.75 ±0.04	+6.5% (noise)	0.047
`poi_reorient`	0.50 ±0.02	0.47 ±0.02	-6.4% (noise)	0.019
`nii_extract_label`	0.49 ±0.02	0.46 ±0.02	-6.2% (noise)	0.051
`nii_rescale_seg`	0.84 ±0.03	0.89 ±0.02	+6.1% (noise)	0.025
`nii_fill_holes`	0.93 ±0.01	0.98 ±0.04	+5.6% (noise)	0.066
`poi_to_global`	0.25 ±0.09	0.24 ±0.02	-5.1% (noise)	0.341
`nii_save`	1.54 ±0.05	1.46 ±0.02	-5.1% (noise)	0.004
`nii_set_dtype`	0.46 ±0.02	0.48 ±0.02	+5.1% (noise)	0.345
`poi_local_to_global_arr`	0.34 ±0.01	0.32 ±0.02	-4.0% (noise)	0.425
`nii_erode_msk`	1.10 ±0.06	1.07 ±0.02	-2.6% (noise)	0.154
`poi_calc_centroids`	1.00 ±0.02	0.98 ±0.01	-2.5% (noise)	0.061
`nii_compute_crop`	0.15 ±0.01	0.15 ±0.01	+1.8% (noise)	0.844
`nii_filter_connected_components`	0.89 ±0.02	0.91 ±0.19	+1.6% (noise)	0.301
`nii_load_seg`	1.82 ±0.38	1.85 ±0.20	+1.6% (noise)	0.684
`poi_rescale`	0.26 ±0.01	0.26 ±0.02	-1.4% (noise)	0.986
`nii_reorient`	0.90 ±0.09	0.89 ±0.04	-1.4% (noise)	0.299
`nii_map_labels`	0.54 ±0.08	0.53 ±0.02	-1.1% (noise)	0.437
`poi_calc_centroids_nocrop`	1.22 ±0.24	1.22 ±0.02	+0.6% (noise)	0.442
`nii_pad_to`	0.61 ±0.04	0.61 ±0.01	+0.4% (noise)	0.264
`nii_load_img`	1.63 ±0.03	1.63 ±0.03	+0.4% (noise)	0.481
`poi_load`	1.11 ±0.08	1.11 ±0.06	-0.1% (noise)	0.290
`nii_volumes`	0.31 ±0.02	0.31 ±0.01	+0.1% (noise)	0.993
`poi_resample_from_to`	0.79 ±0.03	0.79 ±0.02	+0.1% (noise)	0.857

`mri_3d` — shape (68, 52, 67)

Measurement	baseline ms (median ±½·range)	head ms (median ±½·range)	Δ %	p
`nii_fill_holes`	5.41 ±0.09	5.66 ±0.08	+4.6%	0.002
`poi_calc_poi_from_subreg_vert`	11.97 ±0.06	12.40 ±0.08	+3.6%	0.000
`nii_filter_connected_components`	4.70 ±0.10	4.87 ±0.07	+3.6%	0.005
`nii_erode_msk`	7.80 ±0.06	8.01 ±0.27	+2.6%	0.055
`poi_calc_centroids_nocrop`	8.93 ±0.17	9.07 ±0.15	+1.6%	0.393

… 31 more measurements

Measurement	baseline ms (median ±½·range)	head ms (median ±½·range)	Δ %	p
`nii_load_img`	8.86 ±0.14	8.74 ±0.11	-1.4%	0.074
`nii_rescale_seg`	3.42 ±0.09	3.39 ±0.12	-1.0%	0.409
`nii_save`	9.41 ±0.10	9.35 ±0.10	-0.6%	0.298
`nii_rescale`	38.09 ±0.42	37.92 ±0.28	-0.4%	0.266
`nii_resample_from_to`	40.26 ±0.41	40.11 ±0.45	-0.4%	0.522
`nii_dilate_msk`	164.24 ±0.28	164.03 ±1.97	-0.1%	0.833
`metric_voxels`	236912.00 ±0.00	236912.00 ±0.00	+0.0%	—
`metric_labels`	8.00 ±0.00	8.00 ±0.00	+0.0%	—
`metric_foreground_pct`	18.21 ±0.00	18.21 ±0.00	+0.0%	—
`nii_extract_label`	0.51 ±0.02	0.45 ±0.02	-11.4% (noise)	0.005
`nii_get_array`	0.05 ±0.00	0.06 ±0.00	+10.7% (noise)	0.003
`nii_set_dtype`	0.58 ±0.06	0.53 ±0.01	-9.1% (noise)	0.071
`nii_map_labels`	1.12 ±0.06	1.03 ±0.05	-8.0% (noise)	0.021
`poi_rescale`	0.31 ±0.01	0.29 ±0.01	-6.9% (noise)	0.006
`poi_to_global`	0.27 ±0.02	0.26 ±0.02	-4.6% (noise)	0.503
`poi_load`	2.46 ±0.15	2.35 ±0.31	-4.4% (noise)	0.580
`poi_calc_centroids`	2.81 ±0.06	2.93 ±0.08	+4.2% (noise)	0.068
`nii_unique`	0.70 ±0.02	0.67 ±0.01	-4.2% (noise)	0.017
`nii_compute_crop`	0.34 ±0.03	0.33 ±0.01	-3.5% (noise)	0.367
`nii_reorient`	0.94 ±0.03	0.91 ±0.05	-3.4% (noise)	0.307
`poi_resample_from_to`	0.79 ±0.02	0.81 ±0.02	+2.1% (noise)	0.199
`poi_save`	0.57 ±0.02	0.56 ±0.02	-2.0% (noise)	0.357
`poi_map_labels`	0.25 ±0.02	0.25 ±0.02	-1.8% (noise)	0.887
`nii_pad_to`	0.76 ±0.04	0.75 ±0.04	-1.5% (noise)	0.748
`nii_center_of_masses`	2.01 ±0.08	1.99 ±0.02	-1.3% (noise)	0.195
`nii_apply_crop`	0.72 ±0.02	0.71 ±0.01	-0.9% (noise)	0.511
`nii_get_connected_components`	2.36 ±0.17	2.38 ±0.12	+0.9% (noise)	0.684
`poi_local_to_global_arr`	0.35 ±0.02	0.35 ±0.01	-0.7% (noise)	0.315
`poi_reorient`	0.48 ±0.01	0.48 ±0.03	+0.3% (noise)	0.871
`nii_volumes`	1.97 ±0.03	1.98 ±0.02	+0.3% (noise)	0.910
`nii_load_seg`	2.91 ±0.07	2.91 ±0.09	-0.1% (noise)	0.553

`mri_2d` — shape (68, 52, 1)

Measurement	baseline ms (median ±½·range)	head ms (median ±½·range)	Δ %	p
`nii_dilate_msk`	19.64 ±0.09	20.20 ±0.05	+2.8%	0.000
`nii_resample_from_to`	7.60 ±0.09	7.65 ±0.07	+0.7%	0.174
`nii_rescale`	5.48 ±0.16	5.46 ±0.05	-0.4%	0.666
`metric_voxels`	3536.00 ±0.00	3536.00 ±0.00	+0.0%	—
`metric_labels`	7.00 ±0.00	7.00 ±0.00	+0.0%	—

… 30 more measurements

Measurement	baseline ms (median ±½·range)	head ms (median ±½·range)	Δ %	p
`metric_foreground_pct`	50.17 ±0.00	50.17 ±0.00	+0.0%	—
`nii_compute_crop`	0.14 ±0.00	0.16 ±0.01	+12.5% (noise)	0.004
`poi_rescale`	0.28 ±0.01	0.32 ±0.04	+11.7% (noise)	0.113
`poi_resample_from_to`	0.80 ±0.02	0.89 ±0.01	+11.4% (noise)	0.000
`nii_pad_to`	0.54 ±0.02	0.59 ±0.04	+9.3% (noise)	0.095
`poi_local_to_global_arr`	0.36 ±0.02	0.33 ±0.01	-8.0% (noise)	0.031
`poi_save`	0.57 ±0.03	0.61 ±0.04	+7.5% (noise)	0.510
`nii_rescale_seg`	0.78 ±0.03	0.83 ±0.09	+7.3% (noise)	0.226
`nii_unique`	0.12 ±0.01	0.13 ±0.01	+7.2% (noise)	0.085
`nii_save`	1.51 ±0.07	1.40 ±0.02	-6.8% (noise)	0.036
`nii_extract_label`	0.42 ±0.02	0.45 ±0.02	+6.4% (noise)	0.170
`poi_calc_centroids`	1.19 ±0.10	1.25 ±0.21	+4.8% (noise)	0.405
`poi_calc_centroids_nocrop`	1.45 ±0.03	1.52 ±0.02	+4.4% (noise)	0.017
`nii_erode_msk`	1.34 ±0.01	1.40 ±0.02	+4.3% (noise)	0.000
`nii_fill_holes`	1.16 ±0.03	1.19 ±0.23	+2.9% (noise)	0.256
`poi_to_global`	0.27 ±0.03	0.27 ±0.00	+2.9% (noise)	0.765
`nii_get_connected_components`	0.69 ±0.04	0.67 ±0.02	-2.8% (noise)	0.182
`nii_map_labels`	0.55 ±0.03	0.53 ±0.01	-2.5% (noise)	0.725
`nii_filter_connected_components`	0.89 ±0.04	0.91 ±0.01	+2.3% (noise)	0.345
`nii_get_array`	0.02 ±0.01	0.02 ±0.00	+2.2% (noise)	0.990
`nii_load_img`	1.62 ±0.10	1.59 ±0.09	-2.1% (noise)	0.289
`poi_reorient`	0.48 ±0.02	0.49 ±0.02	+1.5% (noise)	0.307
`nii_set_dtype`	0.42 ±0.02	0.43 ±0.01	+1.3% (noise)	0.759
`nii_load_seg`	1.78 ±0.06	1.80 ±0.23	+1.1% (noise)	0.397
`nii_reorient`	0.87 ±0.05	0.88 ±0.01	+1.0% (noise)	0.520
`nii_apply_crop`	0.67 ±0.04	0.67 ±0.01	+1.0% (noise)	0.710
`nii_volumes`	0.50 ±0.01	0.50 ±0.02	-0.7% (noise)	0.267
`nii_center_of_masses`	0.44 ±0.02	0.44 ±0.01	+0.5% (noise)	0.938
`poi_load`	2.21 ±0.09	2.20 ±0.15	-0.4% (noise)	0.916
`poi_map_labels`	0.24 ±0.01	0.24 ±0.01	+0.2% (noise)	0.678

Gate: a measurement fails when the baseline is ≥ 1 ms, the median grows by ≥ 50%, and Welch's t-test gives p < 0.05. metric_* rows are context only.

Memory (peak RSS growth per call)

baseline 7d95bc9 vs head 7d95bc9 · python 3.11.17 · 5 repeats + 1 warmup · sampler proc-status, isolation fork · measurement floor ≈ 0.90 MiB

Showing the 5 most-changed measurements per case; the rest are collapsed. Rows with a baseline below 3 MiB are tagged (noise) — runner jitter on those swamps any real change.

`ct_3d` — shape (73, 47, 73)

Measurement	baseline MiB (median ±½·range)	head MiB (median ±½·range)	Δ %	p
`poi_resample_from_to`	5.18 ±0.00	5.27 ±0.00	+1.8%	—
`nii_reorient`	5.81 ±0.00	5.91 ±0.00	+1.6%	—
`nii_rescale`	6.07 ±0.00	6.15 ±0.00	+1.3%	—
`nii_resample_from_to`	6.20 ±0.00	6.28 ±0.00	+1.3%	—
`nii_load_seg`	5.91 ±0.00	5.99 ±0.00	+1.3%	—

… 31 more measurements

Measurement	baseline MiB (median ±½·range)	head MiB (median ±½·range)	Δ %	p
`nii_dilate_msk`	5.86 ±0.00	5.93 ±0.00	+1.1%	—
`nii_erode_msk`	5.88 ±0.00	5.95 ±0.00	+1.1%	—
`nii_set_dtype`	5.19 ±0.00	5.24 ±0.00	+1.1%	—
`nii_extract_label`	5.31 ±0.00	5.36 ±0.00	+1.0%	—
`nii_pad_to`	5.37 ±0.00	5.42 ±0.00	+0.9%	—
`nii_map_labels`	5.44 ±0.00	5.49 ±0.00	+0.9%	—
`poi_calc_centroids_nocrop`	5.87 ±0.00	5.93 ±0.00	+0.9%	—
`poi_calc_centroids`	6.00 ±0.00	6.05 ±0.00	+0.9%	—
`nii_load_img`	3.52 ±0.00	3.54 ±0.00	+0.7%	—
`nii_rescale_seg`	5.89 ±0.00	5.93 ±0.00	+0.7%	—
`nii_apply_crop`	5.81 ±0.00	5.84 ±0.00	+0.6%	—
`poi_calc_poi_from_subreg_vert`	7.05 ±0.00	7.09 ±0.00	+0.6%	—
`nii_get_connected_components`	6.56 ±0.00	6.60 ±0.00	+0.5%	—
`nii_filter_connected_components`	6.81 ±0.00	6.85 ±0.00	+0.5%	0.000
`nii_save`	3.70 ±0.00	3.69 ±0.00	-0.3%	—
`poi_to_global`	3.08 ±0.00	3.08 ±0.00	+0.1%	—
`nii_fill_holes`	6.27 ±0.00	6.28 ±0.00	+0.1%	—
`poi_reorient`	3.83 ±0.00	3.83 ±0.00	+0.1%	—
`nii_volumes`	3.52 ±0.00	3.52 ±0.00	+0.0%	—
`nii_center_of_masses`	3.27 ±0.00	3.27 ±0.00	+0.0%	—
`metric_voxels`	250463.00 ±0.00	250463.00 ±0.00	+0.0%	—
`metric_labels`	3.00 ±0.00	3.00 ±0.00	+0.0%	—
`metric_foreground_pct`	14.51 ±0.00	14.51 ±0.00	+0.0%	—
`poi_save`	1.39 ±0.00	1.38 ±0.00	-0.8% (noise)	—
`poi_load`	1.61 ±0.00	1.61 ±0.00	+0.2% (noise)	—
`nii_get_array`	1.64 ±0.00	1.65 ±0.00	+0.2% (noise)	—
`poi_map_labels`	1.77 ±0.00	1.77 ±0.00	+0.2% (noise)	—
`poi_local_to_global_arr`	2.95 ±0.00	2.96 ±0.00	+0.1% (noise)	—
`nii_compute_crop`	2.89 ±0.00	2.89 ±0.00	+0.0% (noise)	—
`nii_unique`	2.96 ±0.00	2.96 ±0.00	+0.0% (noise)	—
`poi_rescale`	2.60 ±0.00	2.60 ±0.00	+0.0% (noise)	—

`ct_2d` — shape (73, 47, 1)

Measurement	baseline MiB (median ±½·range)	head MiB (median ±½·range)	Δ %	p
`poi_resample_from_to`	5.12 ±0.00	5.21 ±0.00	+1.8%	—
`nii_reorient`	5.75 ±0.00	5.84 ±0.00	+1.6%	—
`nii_rescale`	6.01 ±0.00	6.09 ±0.00	+1.3%	—
`nii_resample_from_to`	6.13 ±0.00	6.21 ±0.00	+1.3%	—
`nii_load_seg`	5.85 ±0.00	5.92 ±0.00	+1.3%	—

… 30 more measurements

Measurement	baseline MiB (median ±½·range)	head MiB (median ±½·range)	Δ %	p
`nii_dilate_msk`	5.80 ±0.00	5.86 ±0.00	+1.1%	—
`nii_erode_msk`	5.81 ±0.00	5.88 ±0.00	+1.1%	—
`nii_set_dtype`	5.12 ±0.00	5.18 ±0.00	+1.1%	—
`nii_extract_label`	5.25 ±0.00	5.30 ±0.00	+1.0%	—
`nii_pad_to`	5.30 ±0.00	5.36 ±0.00	+1.0%	—
`nii_map_labels`	5.37 ±0.00	5.43 ±0.00	+1.0%	—
`poi_calc_centroids_nocrop`	5.87 ±0.00	5.92 ±0.00	+0.9%	—
`poi_calc_centroids`	5.93 ±0.00	5.99 ±0.00	+0.9%	—
`nii_apply_crop`	5.74 ±0.00	5.78 ±0.00	+0.7%	—
`nii_load_img`	3.45 ±0.00	3.48 ±0.00	+0.7%	—
`nii_rescale_seg`	5.82 ±0.00	5.86 ±0.00	+0.7%	—
`nii_get_connected_components`	6.50 ±0.00	6.54 ±0.00	+0.6%	—
`nii_filter_connected_components`	6.75 ±0.00	6.79 ±0.00	+0.6%	—
`nii_save`	3.64 ±0.00	3.62 ±0.00	-0.3%	—
`poi_to_global`	3.01 ±0.00	3.02 ±0.00	+0.1%	—
`nii_fill_holes`	6.21 ±0.00	6.21 ±0.00	+0.1%	—
`nii_center_of_masses`	3.20 ±0.00	3.21 ±0.00	+0.1%	—
`nii_volumes`	3.45 ±0.00	3.46 ±0.00	+0.1%	—
`poi_reorient`	3.76 ±0.00	3.77 ±0.00	+0.1%	—
`metric_voxels`	3431.00 ±0.00	3431.00 ±0.00	+0.0%	—
`metric_labels`	3.00 ±0.00	3.00 ±0.00	+0.0%	—
`metric_foreground_pct`	37.60 ±0.00	37.60 ±0.00	+0.0%	—
`poi_save`	1.33 ±0.00	1.32 ±0.00	-0.9% (noise)	—
`poi_load`	1.54 ±0.00	1.55 ±0.00	+0.3% (noise)	—
`nii_get_array`	1.58 ±0.00	1.58 ±0.00	+0.2% (noise)	—
`poi_map_labels`	1.70 ±0.00	1.71 ±0.00	+0.2% (noise)	—
`poi_rescale`	2.54 ±0.00	2.54 ±0.00	+0.2% (noise)	—
`nii_compute_crop`	2.83 ±0.00	2.83 ±0.00	+0.1% (noise)	—
`poi_local_to_global_arr`	2.89 ±0.00	2.89 ±0.00	+0.1% (noise)	—
`nii_unique`	2.89 ±0.00	2.89 ±0.00	+0.1% (noise)	—

`mri_3d` — shape (68, 52, 67)

Measurement	baseline MiB (median ±½·range)	head MiB (median ±½·range)	Δ %	p
`poi_resample_from_to`	5.12 ±0.00	5.21 ±0.00	+1.8%	—
`nii_reorient`	5.75 ±0.00	5.84 ±0.00	+1.6%	—
`nii_rescale`	6.01 ±0.00	6.09 ±0.00	+1.3%	—
`nii_resample_from_to`	6.13 ±0.00	6.21 ±0.00	+1.3%	—
`nii_load_seg`	5.85 ±0.00	5.92 ±0.00	+1.3%	—

… 31 more measurements

Measurement	baseline MiB (median ±½·range)	head MiB (median ±½·range)	Δ %	p
`nii_dilate_msk`	5.80 ±0.00	5.86 ±0.00	+1.1%	—
`nii_erode_msk`	5.81 ±0.00	5.88 ±0.00	+1.1%	—
`nii_set_dtype`	5.18 ±0.00	5.24 ±0.00	+1.1%	—
`nii_extract_label`	5.25 ±0.00	5.30 ±0.00	+1.0%	—
`nii_pad_to`	5.30 ±0.00	5.36 ±0.00	+1.0%	—
`nii_map_labels`	5.37 ±0.00	5.43 ±0.00	+1.0%	—
`poi_calc_centroids_nocrop`	5.80 ±0.00	5.86 ±0.00	+0.9%	—
`poi_calc_centroids`	5.93 ±0.00	5.99 ±0.00	+0.9%	—
`nii_apply_crop`	5.74 ±0.00	5.78 ±0.00	+0.7%	—
`nii_rescale_seg`	5.82 ±0.00	5.86 ±0.00	+0.7%	—
`nii_load_img`	3.52 ±0.00	3.54 ±0.00	+0.7%	—
`nii_get_connected_components`	6.50 ±0.00	6.54 ±0.00	+0.6%	—
`nii_filter_connected_components`	6.75 ±0.00	6.79 ±0.00	+0.6%	—
`poi_calc_poi_from_subreg_vert`	7.05 ±0.00	7.09 ±0.00	+0.6%	—
`nii_save`	3.64 ±0.00	3.62 ±0.00	-0.3%	—
`poi_to_global`	3.01 ±0.00	3.02 ±0.00	+0.1%	—
`nii_fill_holes`	6.21 ±0.00	6.21 ±0.00	+0.1%	—
`nii_center_of_masses`	3.20 ±0.00	3.21 ±0.00	+0.1%	—
`nii_volumes`	3.45 ±0.00	3.46 ±0.00	+0.1%	—
`poi_reorient`	3.76 ±0.00	3.77 ±0.00	+0.1%	—
`metric_voxels`	236912.00 ±0.00	236912.00 ±0.00	+0.0%	—
`metric_labels`	8.00 ±0.00	8.00 ±0.00	+0.0%	—
`metric_foreground_pct`	18.21 ±0.00	18.21 ±0.00	+0.0%	—
`poi_save`	1.39 ±0.00	1.38 ±0.00	-0.8% (noise)	—
`poi_load`	1.54 ±0.00	1.55 ±0.00	+0.3% (noise)	—
`nii_get_array`	1.58 ±0.00	1.58 ±0.00	+0.2% (noise)	—
`poi_map_labels`	1.70 ±0.00	1.71 ±0.00	+0.2% (noise)	—
`poi_rescale`	2.54 ±0.00	2.54 ±0.00	+0.2% (noise)	—
`nii_compute_crop`	2.83 ±0.00	2.83 ±0.00	+0.1% (noise)	—
`poi_local_to_global_arr`	2.89 ±0.00	2.89 ±0.00	+0.1% (noise)	—
`nii_unique`	2.89 ±0.00	2.89 ±0.00	+0.1% (noise)	—

`mri_2d` — shape (68, 52, 1)

Measurement	baseline MiB (median ±½·range)	head MiB (median ±½·range)	Δ %	p
`poi_resample_from_to`	5.12 ±0.00	5.21 ±0.00	+1.8%	—
`nii_reorient`	5.75 ±0.00	5.84 ±0.00	+1.6%	—
`nii_rescale`	6.01 ±0.00	6.09 ±0.00	+1.3%	—
`nii_resample_from_to`	6.13 ±0.00	6.21 ±0.00	+1.3%	—
`nii_load_seg`	5.85 ±0.00	5.92 ±0.00	+1.3%	—

… 30 more measurements

Measurement	baseline MiB (median ±½·range)	head MiB (median ±½·range)	Δ %	p
`nii_dilate_msk`	5.80 ±0.00	5.86 ±0.00	+1.1%	—
`nii_erode_msk`	5.81 ±0.00	5.88 ±0.00	+1.1%	—
`nii_set_dtype`	5.18 ±0.00	5.24 ±0.00	+1.1%	—
`nii_extract_label`	5.25 ±0.00	5.30 ±0.00	+1.0%	—
`nii_pad_to`	5.30 ±0.00	5.36 ±0.00	+1.0%	—
`nii_map_labels`	5.37 ±0.00	5.43 ±0.00	+1.0%	—
`poi_calc_centroids_nocrop`	5.87 ±0.00	5.92 ±0.00	+0.9%	—
`poi_calc_centroids`	5.93 ±0.00	5.99 ±0.00	+0.9%	—
`nii_apply_crop`	5.74 ±0.00	5.78 ±0.00	+0.7%	—
`nii_rescale_seg`	5.82 ±0.00	5.86 ±0.00	+0.7%	—
`nii_load_img`	3.52 ±0.00	3.54 ±0.00	+0.7%	—
`nii_get_connected_components`	6.50 ±0.00	6.54 ±0.00	+0.6%	—
`nii_filter_connected_components`	6.75 ±0.00	6.79 ±0.00	+0.6%	—
`nii_save`	3.64 ±0.00	3.62 ±0.00	-0.3%	—
`poi_reorient`	3.76 ±0.00	3.77 ±0.00	+0.2%	—
`poi_to_global`	3.01 ±0.00	3.02 ±0.00	+0.1%	—
`nii_fill_holes`	6.21 ±0.00	6.21 ±0.00	+0.1%	—
`nii_center_of_masses`	3.20 ±0.00	3.21 ±0.00	+0.1%	—
`nii_volumes`	3.45 ±0.00	3.46 ±0.00	+0.1%	—
`metric_voxels`	3536.00 ±0.00	3536.00 ±0.00	+0.0%	—
`metric_labels`	7.00 ±0.00	7.00 ±0.00	+0.0%	—
`metric_foreground_pct`	50.17 ±0.00	50.17 ±0.00	+0.0%	—
`poi_save`	1.39 ±0.00	1.38 ±0.00	-0.8% (noise)	—
`poi_load`	1.54 ±0.00	1.55 ±0.00	+0.3% (noise)	—
`nii_get_array`	1.58 ±0.00	1.58 ±0.00	+0.2% (noise)	—
`poi_map_labels`	1.70 ±0.00	1.71 ±0.00	+0.2% (noise)	—
`poi_rescale`	2.54 ±0.00	2.54 ±0.00	+0.2% (noise)	—
`nii_compute_crop`	2.83 ±0.00	2.83 ±0.00	+0.1% (noise)	—
`poi_local_to_global_arr`	2.89 ±0.00	2.89 ±0.00	+0.1% (noise)	—
`nii_unique`	2.89 ±0.00	2.89 ±0.00	+0.1% (noise)	—

Gate: a measurement fails when the baseline is ≥ 1 MiB, the median grows by ≥ 50%, and Welch's t-test gives p < 0.05. metric_* rows are context only.

Cherry-picked content from tanja@50c8b4d without the whitespace-only reformatting noise: - Add h2d/tse2d1-4/tser2d regex to series-description → format mapping; drop the broad ".*t2w?_tse.*" match. - Fall back to SequenceName when SeriesDescription is empty. Co-Authored-By: Tanja Lerchl <tanja.lerchl@tum.de> Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

…cobb-endplate-v3

Manual fixes in ``TPTBox/spine/spinestats/_load_nako_wh.py``: - D205 (verify_missing_images, verify_hardlink): reword docstrings so the summary line stands alone from the description body. - PLW2901 (``v = preferred`` at line 787): add ``# noqa: PLW2901`` to match the existing noqa on the sibling ``k = mapping.get(k, k)`` rebinding — the loop deliberately overrides both bindings. - PERF102 (line 1385): iterate the ``t2w_chunk`` mapping via ``.values()`` since only the value is used. - TRY300 (``_grid_worker``): move the success ``return`` out of the ``try`` block; only ``_add_grid_info_to_json`` may raise. Everything else is ``ruff check --fix`` and ``ruff format`` output. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

…bb-endplate-v3

robert-graf and others added 25 commits August 26, 2026 14:42

add claude reg info

7f04f69

Merge branch 'Point-Reg' of github.com:Hendrik-code/TPTBox into Point…

d7ef944

…-Reg

Merge branch 'main' of github.com:Hendrik-code/TPTBox

3752050

add smauglab support for internal trainier

d7268a6

edgecases

ea7de02

speed up, split vert and ivd

d0ebb89

update Readme

b12cf44

update nako loader

69fdf16

nako cannonical

4bb6714

add prelimenary scoliose and pelic paramerters

bd99bb6

add NAKO Head scanns

ba8269e

Merge branch 'Point-Reg' of github.com:Hendrik-code/TPTBox into Point…

173af27

…-Reg

Merge branch 'nako-follow-up-export' into Point-Reg

c75ecce

more auto accepts

94bb5d6

add versioning

afe8329

add Verida lordosie version

9467429

add S1 Endplate

06a5f13

add warning to not use is_ct in "DAExt"

840e353

claude_remapping

431a320

fix angls.

7d5b1a5

logging

cfa270a

robert-graf and others added 3 commits September 22, 2026 11:10

style fixes by ruff

2a51ed6

Merge branch 'tanja2' of github.com:Hendrik-code/TPTBox into cleanup/…

a134232

…cobb-endplate-v3

robert-graf added the enhancement New feature or request label Sep 22, 2026

robert-graf and others added 6 commits September 22, 2026 11:18

Merge remote-tracking branch 'origin/poi-label-names' into cleanup/co…

3a08b81

…bb-endplate-v3

fix example test case compile

1a723a8

Merge branch 'main' into cleanup/cobb-endplate-v3

662afff

skipt T13 for apex

cbf9afa

last_vert must be in poi

d58bfd0

robert-graf mentioned this pull request Sep 25, 2026

Stitching refactor + RAM-saving overlap + c_val option + trivial-memory warning #157

Open

5 tasks

robert-graf self-assigned this Oct 1, 2026

Merge branch 'main' into cleanup/cobb-endplate-v3

3b8c9c7

Navigation Menu

Repository navigation

spinestats: version 3 — endplate-plane Cobb + apex + tightened multi-Cobb - #154

spinestats: version 3 — endplate-plane Cobb + apex + tightened multi-Cobb#154
robert-graf wants to merge 35 commits into
mainfrom
cleanup/cobb-endplate-v3

robert-graf commented Sep 22, 2026

Uh oh!

github-actions Bot commented Sep 22, 2026 •

edited

Loading

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Conversation

robert-graf commented Sep 22, 2026

Summary

Changes

Test plan

Uh oh!

github-actions Bot commented Sep 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Benchmark comparison

Speed (wall time per call)

ct_3d — shape (73, 47, 73)

ct_2d — shape (73, 47, 1)

mri_3d — shape (68, 52, 67)

mri_2d — shape (68, 52, 1)

Memory (peak RSS growth per call)

ct_3d — shape (73, 47, 73)

ct_2d — shape (73, 47, 1)

mri_3d — shape (68, 52, 67)

mri_2d — shape (68, 52, 1)

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

github-actions Bot commented Sep 22, 2026 •

edited

Loading

`ct_3d` — shape (73, 47, 73)

`ct_2d` — shape (73, 47, 1)

`mri_3d` — shape (68, 52, 67)

`mri_2d` — shape (68, 52, 1)

`ct_3d` — shape (73, 47, 73)

`ct_2d` — shape (73, 47, 1)

`mri_3d` — shape (68, 52, 67)

`mri_2d` — shape (68, 52, 1)