Hein Dennis, Holmin Staffan, Szczykutowicz Timothy, Maltz Jonathan S, Danielsson Mats, Wang Ge, Persson Mats
Department of Physics, KTH Royal Institute of Technology, Stockholm, 1142, Sweden.
MedTechLabs, Karolinska University Hospital, Stockholm, 17164, Sweden.
Vis Comput Ind Biomed Art. 2024 Sep 23;7(1):24. doi: 10.1186/s42492-024-00175-6.
Deep learning (DL) has proven to be important for computed tomography (CT) image denoising. However, such models are usually trained under supervision, requiring paired data that may be difficult to obtain in practice. Diffusion models offer unsupervised means of solving a wide range of inverse problems via posterior sampling. In particular, using the estimated unconditional score function of the prior distribution, obtained via unsupervised learning, one can sample from the desired posterior via hijacking and regularization. However, due to the iterative solvers used, the number of function evaluations (NFE) required may be orders of magnitudes larger than for single-step samplers. In this paper, we present a novel image denoising technique for photon-counting CT by extending the unsupervised approach to inverse problem solving to the case of Poisson flow generative models (PFGM)++. By hijacking and regularizing the sampling process we obtain a single-step sampler, that is NFE = 1. Our proposed method incorporates posterior sampling using diffusion models as a special case. We demonstrate that the added robustness afforded by the PFGM++ framework yields significant performance gains. Our results indicate competitive performance compared to popular supervised, including state-of-the-art diffusion-style models with NFE = 1 (consistency models), unsupervised, and non-DL-based image denoising techniques, on clinical low-dose CT data and clinical images from a prototype photon-counting CT system developed by GE HealthCare.
深度学习（DL）已被证明对计算机断层扫描（CT）图像去噪很重要。然而，此类模型通常是在监督下训练的，需要成对的数据，而在实际中可能难以获得。扩散模型提供了通过后验采样解决各种逆问题的无监督方法。特别是，利用通过无监督学习获得的先验分布的估计无条件得分函数，人们可以通过劫持和正则化从所需的后验中进行采样。然而，由于使用了迭代求解器，所需的函数评估次数（NFE）可能比单步采样器大几个数量级。在本文中，我们通过将解决逆问题的无监督方法扩展到泊松流生成模型（PFGM）++的情况，提出了一种用于光子计数CT的新型图像去噪技术。通过劫持和正则化采样过程，我们获得了一个单步采样器，即NFE = 1。我们提出的方法将使用扩散模型的后验采样作为一种特殊情况纳入其中。我们证明，PFGM++框架提供的额外鲁棒性带来了显著的性能提升。我们的结果表明，在GE医疗集团开发的原型光子计数CT系统的临床低剂量CT数据和临床图像上，与流行的监督方法（包括具有NFE = 1的先进扩散式模型（一致性模型））、无监督方法以及基于非DL的图像去噪技术相比，我们的方法具有竞争力。
