Improving detection accuracy of perfusion defect in standard dose SPECT-myocardial perfusion imaging by deep-learning denoising.

Improving detection accuracy of perfusion defect in standard dose SPECT-myocardial perfusion imaging by deep-learning denoising.
复制标题

DOI:
10.1007/s12350-021-02676-w
复制
发表时间:
2022-10
期刊:
Journal of nuclear cardiology : official publication of the American Society of Nuclear Cardiology
影响因子:
--
通讯作者:
--
中科院分区:
其他
文献类型:
--
作者:

文献摘要

参考文献

相似文献

我们之前开发了一种深度学习 (DL) 网络,用于 SPECT 心肌灌注成像 (MPI) 中的图像去噪。在这里,我们研究该 DL 网络是否可用于改进标准剂量临床采集中灌注缺陷的检测。为了量化灌注缺陷检测的准确性,我们使用来自 190 名受试者的一组临床 SPECT-MPI 数据,对经过和未经 DL 网络处理的重建图像进行了接受者操作特征 (ROC) 分析。对于灌注缺陷检测,混合研究被用作基本事实,这些研究是根据插入模拟真实病变的临床正常研究创建的。我们考虑了有序子集期望最大化 (OSEM) 重建,其中包括衰减、分辨率和散射校正以及 3D 高斯后滤波。通过定量灌注 SPECT (QPS) 软件计算的总灌注不足 (TPD) 评分用于评估重建图像。与采用最佳高斯后置滤波(sigma=1.2 体素)的重建相比,进一步的 DL 去噪将 ROC 曲线下面积 (AUC) 从 0.80 增加到 0.88(p 值 < 10−4)。对于使用较少高斯后滤波(sigma=0.8 体素)的重建,从而获得更好的空间分辨率,DL 去噪将 AUC 值从 0.78 增加到 0.86(p 值 < 10−4),并在重建中实现了更好的空间分辨率。与传统重建相比,DL 去噪可以有效改善标准剂量 SPECT-MPI 图像中异常缺陷的检测。
We previously developed a deep-learning (DL) network for image denoising in SPECT-myocardial perfusion imaging (MPI). Here we investigate whether this DL network can be utilized for improving detection of perfusion defects in standard-dose clinical acquisitions. To quantify perfusion-defect detection accuracy, we conducted a receiver-operating characteristic (ROC) analysis on reconstructed images with and without processing by the DL network using a set of clinical SPECT-MPI data from 190 subjects. For perfusion-defect detection hybrid studies were used as ground truth, which were created from clinically normal studies with simulated realistic lesions inserted. We considered ordered-subsets expectation-maximization (OSEM) reconstruction with corrections for attenuation, resolution, and scatter and with 3D Gaussian post-filtering. Total perfusion deficit (TPD) scores, computed by Quantitative Perfusion SPECT (QPS) software, were used to evaluate the reconstructed images. Compared to reconstruction with optimal Gaussian post-filtering (sigma=1.2 voxels), further DL denoising increased the area-under-the-ROC-curve (AUC) from 0.80 to 0.88 (p-value < 10−4). For reconstruction with less Gaussian post-filtering (sigma=0.8 voxels), thus better spatial resolution, DL denoising increased the AUC value from 0.78 to 0.86 (p-value < 10−4) and achieved better spatial resolution in reconstruction. DL denoising can effectively improve the detection of abnormal defects in standard-dose SPECT-MPI images over conventional reconstruction.
DOI: 10.1016/j.nuclcard.2004.10.006
发表时间: 2005-01-01
影响因子: 2.4
作者:
Slomka, PJ;Nishina, H;Germano, G
通讯作者: Germano, G
DOI: 10.1109/42.97591
发表时间: 1991-09-01
影响因子: 10.6
作者:
OGAWA, K;HARATA, Y;HASHIMOTO, S
通讯作者: HASHIMOTO, S
使用加权高阶奇异值分解对 3D PET 图像进行重建后滤波
DOI: 10.1186/s12938-016-0221-y
发表时间: 2016-08-27
影响因子: 3.9
作者:
Liu H;Wang K;Tian J
通讯作者: Tian J
DOI: 10.2967/jnumed.112.108969
发表时间: 2013-02-01
影响因子: 9.3
作者:
Arsanjani, Reza;Xu, Yuan;Slomka, Piotr
通讯作者: Slomka, Piotr
DOI: 10.1007/s12350-017-0920-1
发表时间: 2018-12
期刊: Journal of nuclear cardiology : official publication of the American Society of Nuclear Cardiology
影响因子: --
作者:
通讯作者: --