DEep learning-based rapid Spiral Image REconstruction (DESIRE) for high-resolution spiral first-pass myocardial perfusion imaging.
DEep learning-based rapid Spiral Image REconstruction (DESIRE) for high-resolution spiral first-pass myocardial perfusion imaging.
复制标题
基于深度学习的快速螺旋图像重建（DESIRE），用于高分辨率螺旋首过心肌灌注成像。
DOI：
10.1002/nbm.4661
复制
发表时间：
2022
影响因子：
2.9
通讯作者：
Salerno,Michael
中科院分区：
文献类型：
--
作者：
Wang,Junyu;Weller,DanielS;Kramer,ChristopherM;Salerno,Michael
本研究的目的是开发和评价一种基于深度学习的快速螺旋图像重建（DEEP RE）技术，用于全心脏覆盖的高分辨率螺旋首过心肌灌注成像，为单层（SS）和同步多层（SMS）采集提供快速准确的图像重建。在3 T下评价了基于三维U‐ Net的图像增强架构的高分辨率螺旋灌注成像。在来自20名受试者的156个切片的SS灌注图像上训练SS和SMS MB = 2网络。评估了结构相似性指数（SSIM）、峰值信噪比（PSNR）和归一化均方根误差（NRMSE），并由两名经验丰富的心脏病专家对前瞻性图像进行盲法分级（5：优; 1：差）。所提出的技术表现出优异的性能。对于SS，对于最佳网络，SSIM、PSNR和NRMSE分别为0.977 [0.972，0.982]、42.113 [40.174，43.493] dB和0.102 [0.080，0.125]。对于SMS MB = 2的回顾性数据，最佳网络的SSIM、PSNR和NRMSE分别为0.961 [0.950，0.969]、40.834 [39.619，42.004] dB和0.107 [0.086，0.133]。SS L1‐SPIRiT、SS L1‐SPIRiT、MB = 2 SMS‐slice‐L1 ‐ SPIRiT的图像质量评分分别为4.5 [4.1，4.8]、4.5 [4.3，4.6]、3.5 [3.3，4]和3.5 [3.3，3.8]，显示无统计学显著差异（SS和SMS分别为p= 1和p = 1）。每个动态灌注序列的网络推理时间约为100 ms，而L1‐SPIRiT的GPU加速重建时间约为30 min。得出的结论是，对于SS和SMS MB = 2的全心脏高分辨率螺旋灌注成像，RESPONRE能够实现快速和高质量的图像重建。
The objective of the current study was to develop and evaluate a DEep learning‐based rapid Spiral Image REconstruction (DESIRE) technique for high‐resolution spiral first‐pass myocardial perfusion imaging with whole‐heart coverage, to provide fast and accurate image reconstruction for both single‐slice (SS) and simultaneous multislice (SMS) acquisitions. Three‐dimensional U‐Net–based image enhancement architectures were evaluated for high‐resolution spiral perfusion imaging at 3 T. The SS and SMS MB = 2 networks were trained on SS perfusion images from 156 slices from 20 subjects. Structural similarity index (SSIM), peak signal‐to‐noise ratio (PSNR), and normalized root mean square error (NRMSE) were assessed, and prospective images were blindly graded by two experienced cardiologists (5: excellent; 1: poor). Excellent performance was demonstrated for the proposed technique. For SS, SSIM, PSNR, and NRMSE were 0.977 [0.972, 0.982], 42.113 [40.174, 43.493] dB, and 0.102 [0.080, 0.125], respectively, for the best network. For SMS MB = 2 retrospective data, SSIM, PSNR, and NRMSE were 0.961 [0.950, 0.969], 40.834 [39.619, 42.004] dB, and 0.107 [0.086, 0.133], respectively, for the best network. The image quality scores were 4.5 [4.1, 4.8], 4.5 [4.3, 4.6], 3.5 [3.3, 4], and 3.5 [3.3, 3.8] for SS DESIRE, SS L1‐SPIRiT, MB = 2 DESIRE, and MB = 2 SMS‐slice‐L1‐SPIRiT, respectively, showing no statistically significant difference (p= 1 andp= 1 for SS and SMS, respectively) between L1‐SPIRiT and the proposed DESIRE technique. The network inference time was ~100 ms per dynamic perfusion series with DESIRE, while the reconstruction time of L1‐SPIRiT with GPU acceleration was ~ 30 min. It was concluded that DESIRE enabled fast and high‐quality image reconstruction for both SS and SMS MB = 2 whole‐heart high‐resolution spiral perfusion imaging.