Deep learning with noise-to-noise training for denoising in SPECT myocardial perfusion imaging.
Deep learning with noise-to-noise training for denoising in SPECT myocardial perfusion imaging.
复制标题
DOI：
10.1002/mp.14577
复制
发表时间：
2021-01
期刊：
影响因子：
3.8
通讯作者：
King MA
中科院分区：
文献类型：
--
作者：
Liu J;Yang Y;Wernick MN;Pretorius PH;King MA
在单光子发射计算机断层扫描（SPECT）心肌灌注成像（MPI）中，由于数据量有限，通常采用重建后滤波来抑制噪声。我们研究了一种用于传统SPECT-MPI采集中去噪的深度学习（DL）方法，并研究了与传统后滤波相比，它是否能更有效地提高灌注缺陷的可检测性。由于临床研究中缺乏基础事实，我们采用噪声对噪声（N2 N）训练方法对SPECT-MPI图像进行去噪。我们考虑一个耦合的U-网（CU-网）的结构，旨在提高学习效率，通过特征图重用。对于网络训练，我们采用自举程序从列表模式临床采集中生成多个噪声实现。在实验中，我们证明了所提出的方法对一组895临床研究，其中迭代OSEM算法与3D高斯后滤波被用来重建图像。我们研究了使用非预白化匹配滤波器（NPWMF）重建图像中灌注缺陷的检测性能，从图像强度角度评估了左心室（LV）壁的均匀性，并量化了平滑对重建左心室壁空间分辨率的影响。半高宽（FWHM）。与采用高斯后滤波的OSEM相比，采用CU-Net的DL去噪图像在所有对比度水平（65%、50%、35%和20%）下显著提高了灌注缺陷的检测性能。NPWMF输出的信噪比（SNRD）比最佳高斯平滑平均增加8%（p值< 10−4，配对t检验），而受试者间变异性大大降低。CU-Net在SNRD方面也优于3D非局部均值（NLM）滤波器和卷积自动编码器（CAE）去噪网络。此外，重建图像中左室壁的半高宽变化小于1%。此外，CU-Net还提高了检测性能，当图像处理时，重建后平滑较少（为了更好的LV分辨率而增加噪声的权衡），SNRD平均提高了23%。所提出的DL与N2 N训练方法可以产生额外的噪声抑制SPECT-MPI图像比传统的后置滤波。对于灌注缺陷检测，具有CU-Net的DL可以优于具有最佳设置的传统3D高斯滤波以及NLM和CAE。
Post-reconstruction filtering is often applied for noise suppression due to limited data counts in myocardial perfusion imaging (MPI) with single-photon emission computed tomography (SPECT). We study a deep learning (DL) approach for denoising in conventional SPECT-MPI acquisitions, and investigate whether it can be more effective for improving the detectability of perfusion defects compared to traditional post-filtering. Owing to the lack of ground truth in clinical studies, we adopt a noise-to-noise (N2N) training approach for denoising in SPECT-MPI images. We consider a coupled U-Net (CU-Net) structure which is designed to improve learning efficiency through feature-map reuse. For network training we employ a bootstrap procedure to generate multiple noise realizations from list-mode clinical acquisitions. In the experiments we demonstrated the proposed approach on a set of 895 clinical studies, where the iterative OSEM algorithm with 3D Gaussian post-filtering was used to reconstruct the images. We investigated the detection performance of perfusion defects in the reconstructed images using the non-prewhitening matched filter (NPWMF), evaluated the uniformity of left ventricular (LV) wall in terms of image intensity, and quantified the effect of smoothing on the spatial resolution of the reconstructed left ventricular wall by using its full-width at half-maximum (FWHM). Compared to OSEM with Gaussian post-filtering, the DL denoised images with CU-Net significantly improved the detection performance of perfusion defects at all contrast levels (65%, 50%, 35%, and 20%). The signal-to-noise ratio (SNRD) in the NPWMF output was increased on average by 8% over optimal Gaussian smoothing (p-value < 10−4, paired t-test), while the inter-subject variability was greatly reduced. The CU-Net also outperformed a 3D non-local means (NLM) filter and a convolutional autoencoder (CAE) denoising network in terms of SNRD. In addition, the FWHM of the LV wall in the reconstructed images was varied by less than 1%. Furthermore, CU-Net also improved the detection performance when the images were processed with less post-reconstruction smoothing (a trade-off of increased noise for better LV resolution), with SNRD improved on average by 23%. The proposed DL with N2N training approach can yield additional noise suppression in SPECT-MPI images over conventional post-filtering. For perfusion defect detection, DL with CU-Net could outperform conventional 3D Gaussian filtering with optimal setting as well as NLM and CAE.
登录
查看更多内容
影响因子：
3.5
作者：
Buvat, I
通讯作者：
Buvat, I
影响因子：
2.4
作者：
Pretorius, P. Hendrik;Ramon, Albert Juan;Wernick, Miles N.
通讯作者：
Wernick, Miles N.
影响因子：
10.6
作者：
Li, Xiaomeng;Chen, Hao;Heng, Pheng-Ann
通讯作者：
Heng, Pheng-Ann
影响因子：
2.5
作者：
Liu, Junchi;Kocak, Mehmet;Deng, Jie
通讯作者：
Deng, Jie
影响因子：
10.6
作者：
Qi W;Yang Y;Song C;Wernick MN;Pretorius PH;King MA
通讯作者：
King MA