Public

Ultrasound — Public Tier

(3 scenes)

Full-access development tier with all data visible.

Measurements (y), ideal forward operator (H), spec ranges, ground truth (x_true), and true mismatch spec.

Load HDF5 → compare reconstruction vs x_true → check consistency → iterate.

Reconstructed signals (x_hat) and corrected spec as HDF5.

✓

True spec visible — use these exact values for Scenario III oracle reconstruction.

Parameter	Spec Range	True Value	Unit
sos	1520.0 – 1580.0	1550.0	m/s
attenuation	0.4 – 0.7	0.55	dB/cm/MHz
element_sensitivity	-5.0 – 10.0	2.5	%
phase_aberration	-0.3 – 0.6	0.15	rad

Method: CPU_baseline — Mismatch parameter: nominal

11.50 dB

SSIM 0.0990

11.07 dB

SSIM 0.1317

11.54 dB

SSIM 0.1014

Per-scene breakdown (4 scenes)

Scene	PSNR I	SSIM I	PSNR II	SSIM II	PSNR III	SSIM III
scene_00	10.19	0.0676	9.89	0.0865	10.23	0.0692
scene_01	8.62	0.0544	8.54	0.0772	8.66	0.0576
scene_02	9.07	0.0584	9.61	0.0783	9.13	0.0593
scene_03	18.12	0.2158	16.23	0.2851	18.13	0.2197

#	Method	Score	PSNR	SSIM	Consistency	Trust	Source
1	DiffUS + gradient	0.817	34.33	0.963	0.94	✓ Certified	Chen et al., NeurIPS 2024
2	ScoreUS + gradient	0.798	33.56	0.957	0.9	✓ Certified	Johnson et al., ECCV 2025
3	Phase-ADMM-Net + gradient	0.794	32.94	0.952	0.92	✓ Certified	Hou et al., IEEE TMI 2022
4	AttentionBeam + gradient	0.787	32.58	0.949	0.91	✓ Certified	Xu et al., ECCV 2024
5	BeamDATA + gradient	0.786	32.32	0.946	0.92	✓ Certified	Smith et al., ICCV 2024
6	BeamFormer + gradient	0.785	32.88	0.951	0.88	✓ Certified	Li et al., ICCV 2024
7	UltrasoundFormer + gradient	0.780	32.05	0.943	0.91	✓ Certified	Park et al., CVPR 2024
8	MU-Net + gradient	0.780	31.77	0.94	0.93	✓ Certified	Hyun et al., IEEE TUFFC 2022
9	ABLE + gradient	0.732	28.97	0.9	0.91	✓ Certified	Luijten et al., IEEE TMI 2020
10	PnP-ADMM + gradient	0.662	25.52	0.818	0.9	✓ Certified	Goudarzi et al., 2020
11	PW-DAS + gradient	0.658	25.11	0.806	0.92	✓ Certified	Plane wave synthesis baseline
12	PnP-TV + gradient	0.630	24.36	0.781	0.87	✓ Certified	TV regularization for ultrasound
13	DAS-CF + gradient	0.620	24.02	0.769	0.86	✓ Certified	Capon filter, IEEE 1969
14	DAS + gradient	0.574	22.0	0.69	0.89	✓ Certified	Analytical baseline

y H_ideal spec_ranges x_true true_spec

Format: HDF5

Scenes: 3

0.4 × PSNR_norm + 0.4 × SSIM + 0.2 × (1 − ‖y − Ĥx̂‖/‖y‖)

PSNR: 40% SSIM: 40% Consistency: 20%