CT-based synthetic contrast-enhanced dual-energy CT generation using conditional denoising diffusion probabilistic model.

In this study, a deep learning approach utilizing a conditional denoising diffusion probabilistic model (C-DDPM) was developed to create synthetic contrast-enhanced Dual-energy CT (CE-DECT) images from non-contrast single-energy CT (SECT) scans. CE-DECT scans are crucial in producing iodine density maps and delineating targets and organs-at-risk (OAR), which are essential yet often constrained by the limited availability of Dual-energy CT (DECT) scanners during standard CT simulations for radiation therapy planning. To address this challenge, our proposed approach offers a valuable alternative, mitigating the health risks linked to iodinated contrast agents, particularly for those high-risk patients. In this research, imaging data were collected from 130 head-and-neck (HN) cancer patients, who had undergone both non-contrast SECT and CE-DECT scans. The performance of this approach was evaluated using metrics such as Mean Absolute Error (MAE), Structural Similarity Index (SSIM), and Peak Signal-to-Noise Ratio (PSNR). The evaluation demonstrated promising results, with MAE values of 27.37±3.35 Hounsfield Units (HU) for high-energy CT (H-CT) and 24.57±3.35HU for low-energy CT (L-CT), SSIM values of 0.74±0.22 for H-CT and 0.78±0.22 for L-CT, and PSNR values of 18.51±4.55 decibels (dB) for H-CT and 18.91±4.55 dB for L-CT. These metrics highlight the deep learning model's efficacy and its potential to significantly benefit radiation therapy planning by enabling generation of synthetic contrast DECT, even in facilities that lack DECT scanners. Additionally, it offers a safer alternative imaging solution for patients who are unsuitable for iodine contrast imaging, thereby expanding the reach and effectiveness of advanced imaging in cancer treatment planning. .