PET/CT

Positron Emission Tomography / Computed Tomography

Join the Competition Contribute to Benchmark

Standard reconstruction benchmark — forward model perfectly known, no calibration needed. Score = 0.5 × clip((PSNR−15)/30, 0, 1) + 0.5 × SSIM

#	Method	Score	PSNR (dB)	SSIM	Source
🥇	Score-CT Score-CT Song et al., NeurIPS 2024 39.92 dB SSIM 0.984 Checkpoint unavailable	0.907	39.92	0.984	✓ Certified	Song et al., NeurIPS 2024
🥈	DiffusionCT DiffusionCT Kazemi et al., ECCV 2024 39.68 dB SSIM 0.982 Checkpoint unavailable	0.902	39.68	0.982	✓ Certified	Kazemi et al., ECCV 2024
🥉	CTFormer CTFormer Li et al., ICCV 2024 39.45 dB SSIM 0.980 Checkpoint unavailable	0.897	39.45	0.980	✓ Certified	Li et al., ICCV 2024
4	CT-ViT CT-ViT Guo et al., NeurIPS 2024 39.15 dB SSIM 0.978 Checkpoint unavailable	0.891	39.15	0.978	✓ Certified	Guo et al., NeurIPS 2024
5	DOLCE DOLCE Liu et al., ICCV 2023 38.32 dB SSIM 0.971 Checkpoint unavailable	0.874	38.32	0.971	✓ Certified	Liu et al., ICCV 2023
6	DuDoTrans DuDoTrans Wang et al., MLMIR 2022 37.68 dB SSIM 0.962 Checkpoint unavailable	0.859	37.68	0.962	✓ Certified	Wang et al., MLMIR 2022
7	Learned Primal-Dual Learned Primal-Dual Adler & Oktem, IEEE TMI 2018 36.42 dB SSIM 0.947 Checkpoint unavailable	0.831	36.42	0.947	✓ Certified	Adler & Oktem, IEEE TMI 2018
8	FBPConvNet FBPConvNet Jin et al., IEEE TIP 2017 35.81 dB SSIM 0.939 Checkpoint unavailable	0.816	35.81	0.939	✓ Certified	Jin et al., IEEE TIP 2017
9	RED-CNN RED-CNN Chen et al., IEEE TMI 2017 33.56 dB SSIM 0.908 Checkpoint unavailable	0.763	33.56	0.908	✓ Certified	Chen et al., IEEE TMI 2017
10	PnP-DnCNN PnP-DnCNN Zhang et al., 2017 33.45 dB SSIM 0.905 Try in SpecLab →	0.760	33.45	0.905	✓ Certified	Zhang et al., 2017
11	PnP-ADMM PnP-ADMM Venkatakrishnan et al., 2013 32.64 dB SSIM 0.891 Try in SpecLab →	0.740	32.64	0.891	✓ Certified	Venkatakrishnan et al., 2013
12	TV-ADMM TV-ADMM Sidky et al., 2008 30.15 dB SSIM 0.862 Try in SpecLab →	0.683	30.15	0.862	✓ Certified	Sidky et al., 2008
13	FBP FBP Kak & Slaney, 1988 27.38 dB SSIM 0.790 Try in SpecLab →	0.601	27.38	0.790	✓ Certified	Kak & Slaney, 1988

Dataset: PWM Benchmark (13 algorithms)

Blind Reconstruction Challenge — forward model has unknown mismatch, must calibrate from data. Score = 0.4 × PSNR_norm + 0.4 × SSIM + 0.2 × (1 − ‖y − Ĥx̂‖/‖y‖)

#	Method	Overall Score	Public PSNR / SSIM	Dev PSNR / SSIM	Hidden PSNR / SSIM	Trust	Source
🥇	CTFormer + gradient CTFormer + gradient Li et al., ICCV 2024 Score 0.813 Correct & Reconstruct →	0.813	0.859 38.02 dB / 0.982	0.801 35.3 dB / 0.970	0.779 32.81 dB / 0.951	✓ Certified	Li et al., ICCV 2024
🥈	CT-ViT + gradient CT-ViT + gradient Guo et al., NeurIPS 2024 Score 0.792 Correct & Reconstruct →	0.792	0.838 37.21 dB / 0.979	0.800 34.44 dB / 0.964	0.739 30.53 dB / 0.925	✓ Certified	Guo et al., NeurIPS 2024
🥉	DiffusionCT + gradient DiffusionCT + gradient Kazemi et al., ECCV 2024 Score 0.781 Correct & Reconstruct →	0.781	0.863 38.14 dB / 0.983	0.753 30.78 dB / 0.928	0.727 29.89 dB / 0.915	✓ Certified	Kazemi et al., ECCV 2024
4	Score-CT + gradient Score-CT + gradient Song et al., NeurIPS 2024 Score 0.750 Correct & Reconstruct →	0.750	0.846 38.08 dB / 0.982	0.739 29.54 dB / 0.910	0.666 26.32 dB / 0.841	✓ Certified	Song et al., NeurIPS 2024
5	DuDoTrans + gradient DuDoTrans + gradient Wang et al., MLMIR 2022 Score 0.741 Correct & Reconstruct →	0.741	0.818 35.4 dB / 0.970	0.720 28.88 dB / 0.898	0.684 27.98 dB / 0.880	✓ Certified	Wang et al., MLMIR 2022
6	DOLCE + gradient DOLCE + gradient Liu et al., ICCV 2023 Score 0.729 Correct & Reconstruct →	0.729	0.847 37.14 dB / 0.979	0.692 27.32 dB / 0.866	0.647 26.11 dB / 0.835	✓ Certified	Liu et al., ICCV 2023
7	Learned Primal-Dual + gradient Learned Primal-Dual + gradient Adler & Oktem, IEEE TMI 2018 Score 0.696 Correct & Reconstruct →	0.696	0.804 34.43 dB / 0.964	0.698 27.79 dB / 0.876	0.587 23.65 dB / 0.756	✓ Certified	Adler & Oktem, IEEE TMI 2018
8	PnP-DnCNN + gradient PnP-DnCNN + gradient Zhang et al., IEEE TIP 2017 Score 0.690 Correct & Reconstruct →	0.690	0.763 31.61 dB / 0.938	0.692 27.98 dB / 0.880	0.616 24.86 dB / 0.798	✓ Certified	Zhang et al., IEEE TIP 2017
9	FBPConvNet + gradient FBPConvNet + gradient Jin et al., IEEE TIP 2017 Score 0.689 Correct & Reconstruct →	0.689	0.797 34.03 dB / 0.961	0.670 26.55 dB / 0.847	0.599 23.99 dB / 0.768	✓ Certified	Jin et al., IEEE TIP 2017
10	PnP-ADMM + gradient PnP-ADMM + gradient Venkatakrishnan et al., IEEE GlobalSIP 2013 Score 0.682 Correct & Reconstruct →	0.682	0.776 31.63 dB / 0.939	0.672 25.96 dB / 0.831	0.599 24.08 dB / 0.771	✓ Certified	Venkatakrishnan et al., IEEE GlobalSIP 2013
11	TV-ADMM + gradient TV-ADMM + gradient Sidky et al., Phys. Med. Biol. 2008 Score 0.669 Correct & Reconstruct →	0.669	0.708 27.97 dB / 0.880	0.662 25.75 dB / 0.825	0.638 25.43 dB / 0.815	✓ Certified	Sidky et al., Phys. Med. Biol. 2008
12	RED-CNN + gradient RED-CNN + gradient Chen et al., IEEE TMI 2017 Score 0.643 Correct & Reconstruct →	0.643	0.786 32.2 dB / 0.945	0.602 23.24 dB / 0.740	0.540 21.72 dB / 0.678	✓ Certified	Chen et al., IEEE TMI 2017
13	FBP + gradient FBP + gradient Kak & Slaney, IEEE Press 1988 Score 0.640 Correct & Reconstruct →	0.640	0.644 24.72 dB / 0.793	0.668 26.25 dB / 0.839	0.609 24.5 dB / 0.786	✓ Certified	Kak & Slaney, IEEE Press 1988

Complete score requires all 3 tiers (Public + Dev + Hidden).

Join the competition →

Scoring: 0.4 × PSNR_norm + 0.4 × SSIM + 0.2 × (1 − ‖y − Ĥx̂‖/‖y‖) PSNR 40% · SSIM 40% · Consistency 20%

Public 3 scenes

Full-access development tier with all data visible.

What you get & how to use

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.

Public Leaderboard

#	Method	Score	PSNR	SSIM
1	DiffusionCT + gradient	0.863	38.14	0.983
2	CTFormer + gradient	0.859	38.02	0.982
3	DOLCE + gradient	0.847	37.14	0.979
4	Score-CT + gradient	0.846	38.08	0.982
5	CT-ViT + gradient	0.838	37.21	0.979
6	DuDoTrans + gradient	0.818	35.4	0.97
7	Learned Primal-Dual + gradient	0.804	34.43	0.964
8	FBPConvNet + gradient	0.797	34.03	0.961
9	RED-CNN + gradient	0.786	32.2	0.945
10	PnP-ADMM + gradient	0.776	31.63	0.939
11	PnP-DnCNN + gradient	0.763	31.61	0.938
12	TV-ADMM + gradient	0.708	27.97	0.88
13	FBP + gradient	0.644	24.72	0.793

Spec Ranges (3 parameters)

Parameter	Min	Max	Unit
ct_registration_shift	-4.0	8.0	pixels
hu_to_mu_scale	-10.0	20.0	%
scatter_fraction	-0.15	0.3

View Details →

Dev 3 scenes

Blind evaluation tier — no ground truth available.

What you get & how to use

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.

Dev Leaderboard

#	Method	Score	PSNR	SSIM
1	CTFormer + gradient	0.801	35.3	0.97
2	CT-ViT + gradient	0.800	34.44	0.964
3	DiffusionCT + gradient	0.753	30.78	0.928
4	Score-CT + gradient	0.739	29.54	0.91
5	DuDoTrans + gradient	0.720	28.88	0.898
6	Learned Primal-Dual + gradient	0.698	27.79	0.876
7	DOLCE + gradient	0.692	27.32	0.866
8	PnP-DnCNN + gradient	0.692	27.98	0.88
9	PnP-ADMM + gradient	0.672	25.96	0.831
10	FBPConvNet + gradient	0.670	26.55	0.847
11	FBP + gradient	0.668	26.25	0.839
12	TV-ADMM + gradient	0.662	25.75	0.825
13	RED-CNN + gradient	0.602	23.24	0.74

Spec Ranges (3 parameters)

Parameter	Min	Max	Unit
ct_registration_shift	-4.8	7.2	pixels
hu_to_mu_scale	-12.0	18.0	%
scatter_fraction	-0.18	0.27

Submit Reconstruction View Details →

Hidden 3 scenes

Fully blind server-side evaluation — no data download.

What you get & how to use

What you get: No data downloadable. Algorithm runs server-side on hidden measurements.

How to use: Package algorithm as Docker container / Python script. Submit via link.

What to submit: Containerized algorithm accepting y + H, outputting x_hat + corrected spec.

Hidden Leaderboard

#	Method	Score	PSNR	SSIM
1	CTFormer + gradient	0.779	32.81	0.951
2	CT-ViT + gradient	0.739	30.53	0.925
3	DiffusionCT + gradient	0.727	29.89	0.915
4	DuDoTrans + gradient	0.684	27.98	0.88
5	Score-CT + gradient	0.666	26.32	0.841
6	DOLCE + gradient	0.647	26.11	0.835
7	TV-ADMM + gradient	0.638	25.43	0.815
8	PnP-DnCNN + gradient	0.616	24.86	0.798
9	FBP + gradient	0.609	24.5	0.786
10	FBPConvNet + gradient	0.599	23.99	0.768
11	PnP-ADMM + gradient	0.599	24.08	0.771
12	Learned Primal-Dual + gradient	0.587	23.65	0.756
13	RED-CNN + gradient	0.540	21.72	0.678

Spec Ranges (3 parameters)

Parameter	Min	Max	Unit
ct_registration_shift	-2.8	9.2	pixels
hu_to_mu_scale	-7.0	23.0	%
scatter_fraction	-0.105	0.345

Submit Algorithm View Details →

Blind Reconstruction Challenge

Challenge

Given measurements with unknown mismatch and spec ranges (not exact params), reconstruct the original signal. A method must be evaluated on all three tiers for a complete score. Scored on a composite metric: 0.4 × PSNR_norm + 0.4 × SSIM + 0.2 × (1 − ‖y − Ĥx̂‖/‖y‖).

Input

Measurements y, ideal forward model H, spec ranges

Output

Reconstructed signal x̂

Spec DAG — Forward Model Pipeline

R(θ) → D(μ_ct) → Π(LOR) → D(g, η)

R Radon Transform (θ)

D CT Attenuation Map (μ_ct)

Π PET Line of Response (LOR)

D Detector + Noise (g, η)

Mismatch Parameters

Symbol	Parameter	Description	Perturbed
Δs	ct_registration_shift	CT-PET registration error (pixels)	4.0
Δμ	hu_to_mu_scale	HU-to-μ calibration error (%)	10.0
f_s	scatter_fraction	Scatter fraction	0.15

Credits System

40%

Platform Profit Pool

Revenue allocated to benchmark rewards

30%

Winner Share

Top algorithm receives from pool

$100

Min Withdrawal

Minimum payout threshold

Spec Primitives Reference (11 primitives)

P Propagation

Free-space or medium propagation kernel (Fresnel, Rayleigh-Sommerfeld).

M Mask / Modulation

Spatial or spatio-temporal amplitude modulation (coded aperture, SLM pattern).

Π Projection

Geometric projection operator (Radon transform, fan-beam, cone-beam).

F Fourier Sampling

Sampling in the Fourier / k-space domain (MRI, ptychography).

C Convolution

Shift-invariant convolution with a point-spread function (PSF).

Σ Summation / Integration

Summation along a physical dimension (spectral, temporal, angular).

D Detector

Sensor readout with gain g and noise model η (Gaussian, Poisson, mixed).

S Structured Illumination

Patterned illumination (block, Hadamard, random) applied to the scene.

W Wavelength Dispersion

Spectral dispersion element (prism, grating) with shift α and aperture a.

R Rotation / Motion

Sample or gantry rotation (CT, electron tomography).

Λ Wavelength Selection

Spectral filter or monochromator selecting a wavelength band.