Ueda Takahiro, Ohno Yoshiharu, Yamamoto Kaori, Murayama Kazuhiro, Ikedo Masato, Yui Masao, Hanamatsu Satomu, Tanaka Yumi, Obama Yuki, Ikeda Hirotaka, Toyama Hiroshi
From the Department of Radiology (T.U., Y. Ohno, S.H., Y.T., Y. Obama, H.I., H.T.) and Joint Research Laboratory of Advanced Medical Imaging (Y. Ohno, K.M.), Fujita Health University School of Medicine, 1-98 Dengakugakubo, Kutsukake-cho, Toyoake 470-1192, Japan; and Canon Medical Systems Corporation, Otawara, Japan (K.Y., M.I., M.Y.).
Radiology. 2022 May;303(2):373-381. doi: 10.1148/radiol.204097. Epub 2022 Feb 1.
Background Deep learning reconstruction (DLR) may improve image quality. However, its impact on diffusion-weighted imaging (DWI) of the prostate has yet to be assessed. Purpose To determine whether DLR can improve image quality of diffusion-weighted MRI at values ranging from 1000 sec/mm to 5000 sec/mm in patients with prostate cancer. Materials and Methods In this retrospective study, images of the prostate obtained at DWI with a value of 0 sec/mm, DWI with a value of 1000 sec/mm (DWI), DWI with a value of 3000 sec/mm (DWI), and DWI with a value of 5000 sec/mm (DWI) from consecutive patients with biopsy-proven cancer from January to June 2020 were reconstructed with and without DLR. Image quality was assessed using signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) from region-of-interest analysis and qualitatively assessed using a five-point visual scoring system (1 [very poor] to 5 [excellent]) for each high--value DWI sequence with and without DLR. The SNR, CNR, and visual score for DWI with and without DLR were compared with the paired test and the Wilcoxon signed rank test with Bonferroni correction, respectively. Apparent diffusion coefficients (ADCs) from DWI with and without DLR were also compared with the paired test with Bonferroni correction. Results A total of 60 patients (mean age, 67 years; age range, 49-79 years) were analyzed. DWI with DLR showed significantly higher SNRs and CNRs than DWI without DLR ( < .001); for example, with DWI the mean SNR was 38.7 ± 0.6 versus 17.8 ± 0.6, respectively ( < .001), and the mean CNR was 18.4 ± 5.6 versus 7.4 ± 5.6, respectively ( < .001). DWI with DLR also demonstrated higher qualitative image quality than DWI without DLR (mean score: 4.8 ± 0.4 vs 4.0 ± 0.7, respectively, with DWI [ = .001], 3.8 ± 0.7 vs 3.0 ± 0.8 with DWI [ = .002], and 3.1 ± 0.8 vs 2.0 ± 0.9 with DWI [ < .001]). ADCs derived with and without DLR did not differ substantially ( > .99). Conclusion Deep learning reconstruction improves the image quality of diffusion-weighted MRI scans of prostate cancer with no impact on apparent diffusion coefficient quantitation with a 3.0-T MRI system. © RSNA, 2022 . See also the editorial by Turkbey in this issue.
背景 深度学习重建（DLR）可能会提高图像质量。然而，其对前列腺扩散加权成像（DWI）的影响尚未得到评估。目的 确定DLR是否能改善前列腺癌患者在1000秒/毫米²至5000秒/毫米²范围内的扩散加权磁共振成像（MRI）图像质量。材料与方法 在这项回顾性研究中，对2020年1月至6月连续活检证实为癌症的患者，在 值为0秒/毫米²的DWI、 值为1000秒/毫米²的DWI（DWI1000）、 值为3000秒/毫米²的DWI（DWI3000）和 值为5000秒/毫米²的DWI（DWI5000）时获取的前列腺图像进行有或无DLR的重建。使用感兴趣区分析的信噪比（SNR）和对比噪声比（CNR）评估图像质量，并使用五点视觉评分系统（1[非常差]至5[优秀]）对每个有或无DLR的高 值DWI序列进行定性评估。分别采用配对t检验和经Bonferroni校正的Wilcoxon符号秩检验比较有和无DLR的DWI的SNR、CNR和视觉评分。还采用经Bonferroni校正的配对t检验比较有和无DLR的DWI的表观扩散系数（ADC）。结果 共分析了60例患者（平均年龄67岁；年龄范围49 - 79岁）。有DLR的DWI显示出比无DLR的DWI显著更高的SNR和CNR（P <.001）；例如，对于DWI1 +，平均SNR分别为38.7 ± 0.6和17.8 ± 0.6（P <.001），平均CNR分别为18.4 ± 5.6和7.4 ± 5.6（P <.001）。有DLR的DWI在定性图像质量上也高于无DLR的DWI（平均评分：DWI1000分别为4.8 ± 0.4和4.0 ± 0.7，P =.001；DWI3 +分别为3.8 ± 0.7和3.0 ± 0.8，P =.002；DWI5 +分别为3.1 ± 0.8和2.0 ± 0.9，P <.001）。有和无DLR得出的ADC没有实质性差异（P >.99）。结论 深度学习重建可改善前列腺癌扩散加权MRI扫描的图像质量，且对3.0-T MRI系统的表观扩散系数定量无影响。© RSNA，2022 。另见本期Turkbey的社论。
