Uncertainty quantification in deep-learning (DL) based image reconstruction models is critical for reliable clinical decision making based on the reconstructed images. We introduce "NPB-REC", a non-parametric fully Bayesian framework for uncertainty assessment in MRI reconstruction from undersampled "k-space" data. We use Stochastic gradient Langevin dynamics (SGLD) during the training phase to characterize the posterior distribution of the network weights. We demonstrated the added-value of our approach on the multi-coil brain MRI dataset, from the fastmri challenge, in comparison to the baseline E2E-VarNet with and without inference-time dropout. Our experiments show that NPB-REC outperforms the baseline by means of reconstruction accuracy (PSNR and SSIM of $34.55$, $0.908$ vs. $33.08$, $0.897$, $p<0.01$) in high acceleration rates ($R=8$). This is also measured in regions of clinical annotations. More significantly, it provides a more accurate estimate of the uncertainty that correlates with the reconstruction error, compared to the Monte-Carlo inference time Dropout method (Pearson correlation coefficient of $R=0.94$ vs. $R=0.91$). The proposed approach has the potential to facilitate safe utilization of DL based methods for MRI reconstruction from undersampled data. Code and trained models are available in \url{https://github.com/samahkh/NPB-REC}.
翻译:在基于深层学习(DL)图像重建模型中的不确定性量化对于基于重建图像的可靠临床决策至关重要。我们采用“NPB-REC”这一非参数性的完整巴伊西亚框架,从未充分抽样的“k-space”数据中,在MRI重建中进行不确定性评估;我们在培训阶段使用Stochastestec climate Langevin动态(SGLD)来说明网络重量的后端分布。我们展示了我们从快速挑战到基线E2E-VarNet的附加价值{MRI数据集。更准确地估计了NPB-REC在重建精度方面超过了基线(PSNR和SSIM为34.55美元,0.908美元对33.08美元,0.997美元对美元/0.01美元)的高加速率(R=8美元),在临床说明区域中也测量了这一方法。更重要的是,它提供了与经过培训的E2E-VERNet数据库在重建过程中的不确定性,与正在使用的美元-CREMR=Rml)的利用率下,与正在下降的汇率下的数据比。