Public

Mammography — Public Tier

(3 scenes)

Full-access development tier with all data visible.

What you get

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

How to use

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

What to submit

Reconstructed signals (x_hat) and corrected spec as HDF5.

Parameter Specifications

✓

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

Parameter Spec Range True Value Unit
compression -2.0 – 4.0 1.0 mm
anode_angle -0.5 – 1.0 0.25 deg
scatter 0.25 – 0.4 0.325

InverseNet Baseline Scores

Method: CPU_baseline — Mismatch parameter: nominal

Scenario I (Ideal)

21.29 dB

SSIM 0.8581

Scenario II (Mismatch)

18.47 dB

SSIM 0.8462

Scenario III (Oracle)

1.93 dB

SSIM -0.0339

Per-scene breakdown (4 scenes)
Scene PSNR I SSIM I PSNR II SSIM II PSNR III SSIM III
scene_00 18.76 0.8194 16.52 0.8010 2.03 -0.0291
scene_01 17.57 0.7844 12.95 0.7572 2.16 -0.0790
scene_02 23.79 0.9272 25.60 0.9301 1.61 0.0109
scene_03 25.02 0.9012 18.83 0.8967 1.91 -0.0382

Public Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 Score-CT + gradient 0.866 38.45 0.984 0.94 ✓ Certified Song et al., NeurIPS 2024
2 CT-ViT + gradient 0.857 37.86 0.982 0.93 ✓ Certified Guo et al., NeurIPS 2024
3 DOLCE + gradient 0.847 36.83 0.977 0.94 ✓ Certified Liu et al., ICCV 2023
4 DiffusionCT + gradient 0.843 37.53 0.98 0.88 ✓ Certified Kazemi et al., ECCV 2024
5 CTFormer + gradient 0.842 37.57 0.98 0.87 ✓ Certified Li et al., ICCV 2024
6 Learned Primal-Dual + gradient 0.826 35.4 0.97 0.92 ✓ Certified Adler & Oktem, IEEE TMI 2018
7 DuDoTrans + gradient 0.820 35.84 0.973 0.86 ✓ Certified Wang et al., MLMIR 2022
8 FBPConvNet + gradient 0.816 34.44 0.964 0.93 ✓ Certified Jin et al., IEEE TIP 2017
9 PnP-ADMM + gradient 0.771 30.91 0.93 0.95 ✓ Certified Venkatakrishnan et al., IEEE GlobalSIP 2013
10 RED-CNN + gradient 0.762 30.97 0.93 0.9 ✓ Certified Chen et al., IEEE TMI 2017
11 PnP-DnCNN + gradient 0.759 30.94 0.93 0.89 ✓ Certified Zhang et al., IEEE TIP 2017
12 TV-ADMM + gradient 0.731 28.72 0.895 0.93 ✓ Certified Sidky et al., Phys. Med. Biol. 2008
13 FBP + gradient 0.652 25.33 0.812 0.87 ✓ Certified Kak & Slaney, IEEE Press 1988

Visible Data Fields

y H_ideal spec_ranges x_true true_spec

Dataset

Format: HDF5
Scenes: 3

Scoring Formula

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

PSNR: 40% SSIM: 40% Consistency: 20%
Back to Mammography