Dev

DEXA — Dev Tier

(3 scenes)

Blind evaluation tier — no ground truth available.

What you get

Measurements (y), ideal forward operator (H), and spec ranges only.

How to use

Apply your pipeline from the Public tier. Use consistency as self-check.

What to submit

Reconstructed signals and corrected spec. Scored server-side.

Parameter Specifications

🔒

True spec hidden — estimate parameters from spec ranges below.

Parameter Spec Range Unit
energy_offset -1.2 – 1.8 keV
soft_tissue -3.6 – 5.4 %
beam_overlap -0.024 – 0.036

Dev Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 SwinDXA + gradient 0.804 34.8 0.966 0.85 ✓ Certified Liu et al., ICCV 2021 (DEXA adapt.)
2 PhysDXA + gradient 0.768 31.38 0.936 0.9 ✓ Certified Raissi et al., J. Comput. Phys. 2019 (DEXA)
3 DiffusionDXA + gradient 0.747 31.2 0.933 0.81 ✓ Certified Blattmann et al., arXiv 2023 (DEXA adapt.)
4 PnP-DXA + gradient 0.730 29.11 0.902 0.89 ✓ Certified Venkatakrishnan et al., 2013 (DEXA adapt.)
5 DXA-CNN + gradient 0.686 27.48 0.869 0.82 ✓ Certified Lee et al., Bone 2020
6 BML-Sep + gradient 0.658 26.14 0.836 0.81 ✓ Certified Lehmann et al., Med. Phys. 1981
7 DXA-U-Net + gradient 0.651 25.18 0.808 0.88 ✓ Certified Huo et al., IEEE TMED 2021
8 TV-DEXA + gradient 0.604 23.36 0.745 0.86 ✓ Certified Sidky & Pan, Phys. Med. Biol. 2008 (DEXA)
9 FBP-DEXA + gradient 0.600 23.77 0.76 0.79 ✓ Certified Mazess et al., Am. J. Clin. Nutr. 1990

Visible Data Fields

y H_ideal spec_ranges

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 DEXA