What question did this study set out to answer?

To develop a conditional 3D latent diffusion model for synthesizing PET from CT scans in the head–neck and thoracic regions.

May 6, 2026BioengineeringOpen Access

CT-to-PET Synthesis in the Head–Neck and Thoracic Region via Conditional 3D Latent Diffusion Modeling

View Full Paper

Ask AI

Bookmark

Authors

Proposed model improves PET synthesis accuracy in head-neck and thoracic regions, suggesting implications for clinical decision support.

To develop a conditional 3D latent diffusion model for synthesizing PET from CT scans in the head–neck and thoracic regions.
Proposed a conditional 3D latent diffusion framework (3D-LDM) for CT-to-PET synthesis.
Segmented lung anatomy from CT to limit variability in PET generation.
Trained the model on 900 paired PET/CT studies and evaluated performance using MAE, PSNR, and SSIM.
3D-LDM achieved the best quantitative fidelity on the test cohort with MAE = 303.05 ± 22.16 SUV units.
Model precision and recall at lesion level were both 0.76, detecting an average of 3.19 lesions per scan.
False-positive rate was measured at 0.72 per scan, with lesion-wise NMSE at 11.37% outperforming GAN and transformer baselines.

Mahdi et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark

May 6, 2026BioengineeringOpen Access

View Full Paper

Ask AI

Bookmark

Authors

Proposed model improves PET synthesis accuracy in head-neck and thoracic regions, suggesting implications for clinical decision support.

To develop a conditional 3D latent diffusion model for synthesizing PET from CT scans in the head–neck and thoracic regions.
Proposed a conditional 3D latent diffusion framework (3D-LDM) for CT-to-PET synthesis.
Segmented lung anatomy from CT to limit variability in PET generation.
Trained the model on 900 paired PET/CT studies and evaluated performance using MAE, PSNR, and SSIM.
3D-LDM achieved the best quantitative fidelity on the test cohort with MAE = 303.05 ± 22.16 SUV units.
Model precision and recall at lesion level were both 0.76, detecting an average of 3.19 lesions per scan.
False-positive rate was measured at 0.72 per scan, with lesion-wise NMSE at 11.37% outperforming GAN and transformer baselines.

Mahdi et al. (2026) studied this question.

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark