Haase Robert, Pinetz Thomas, Kobler Erich, Bendella Zeynep, Gronemann Christian, Paech Daniel, Radbruch Alexander, Effland Alexander, Deike Katerina
From the Clinic of Neuroradiology, University Hospital Bonn, Rheinische Friedrich-Wilhelms-Universität Bonn, Bonn, Germany (R.H., E.K., Z.B., C.G., D.P., A.R., K.D.); Institute of Applied Mathematics, Rheinische Friedrich-Wilhelms-Universität Bonn, Bonn, Germany (T.P., A.E.); Department of Radiology, Brigham and Women's Hospital, Harvard Medical School, Boston, MA (D.P.); and German Center for Neurodegenerative Diseases (DZNE), Helmholtz Association of German Research Centers, Bonn, Germany (A.R., K.D.).
Invest Radiol. 2025 Feb 1;60(2):105-113. doi: 10.1097/RLI.0000000000001107. Epub 2024 Jul 30.
Reducing gadolinium-based contrast agents to lower costs, the environmental impact of gadolinium-containing wastewater, and patient exposure is still an unresolved issue. Published methods have never been compared. The purpose of this study was to compare the performance of 2 reimplemented state-of-the-art deep learning methods (settings A and B) and a proposed method for contrast signal extraction (setting C) to synthesize artificial T1-weighted full-dose images from corresponding noncontrast and low-dose images.
In this prospective study, 213 participants received magnetic resonance imaging of the brain between August and October 2021 including low-dose (0.02 mmol/kg) and full-dose images (0.1 mmol/kg). Fifty participants were randomly set aside as test set before training (mean age ± SD, 52.6 ± 15.3 years; 30 men). Artificial and true full-dose images were compared using a reader-based study. Two readers noted all false-positive lesions and scored the overall interchangeability in regard to the clinical conclusion. Using a 5-point Likert scale (0 being the worst), they scored the contrast enhancement of each lesion and its conformity to the respective reference in the true image.
The average counts of false-positives per participant were 0.33 ± 0.93, 0.07 ± 0.33, and 0.05 ± 0.22 for settings A-C, respectively. Setting C showed a significantly higher proportion of scans scored as fully or mostly interchangeable (70/100) than settings A (40/100, P < 0.001) and B (57/100, P < 0.001), and generated the smallest mean enhancement reduction of scored lesions (-0.50 ± 0.55) compared with the true images (setting A: -1.10 ± 0.98; setting B: -0.91 ± 0.67, both P < 0.001). The average scores of conformity of the lesion were 1.75 ± 1.07, 2.19 ± 1.04, and 2.48 ± 0.91 for settings A-C, respectively, with significant differences among all settings (all P < 0.001).
The proposed method for contrast signal extraction showed significant improvements in synthesizing postcontrast images. A relevant proportion of images showing inadequate interchangeability with the reference remains at this dosage.
减少基于钆的造影剂以降低成本、含钆废水对环境的影响以及患者暴露量仍是一个未解决的问题。已发表的方法从未被比较过。本研究的目的是比较两种重新实现的先进深度学习方法（设置A和B）和一种提出的造影剂信号提取方法（设置C）从相应的非增强和低剂量图像合成人工T1加权全剂量图像的性能。
在这项前瞻性研究中，213名参与者于2021年8月至10月接受了脑部磁共振成像，包括低剂量（0.02 mmol/kg）和全剂量图像（0.1 mmol/kg）。在训练前随机留出50名参与者作为测试集（平均年龄±标准差，52.6±15.3岁；30名男性）。使用基于阅片者的研究比较人工合成和真实的全剂量图像。两名阅片者记录所有假阳性病变，并就临床结论对整体互换性进行评分。使用5分李克特量表（0为最差），他们对每个病变的对比增强及其与真实图像中相应参考的符合程度进行评分。
设置A - C中，每位参与者的平均假阳性计数分别为0.33±0.93、0.07±0.33和0.05±0.22。设置C显示被评为完全或大部分可互换的扫描比例（70/100）显著高于设置A（40/100，P < 0.001）和设置B（57/100，P < 0.001），并且与真实图像相比，设置C产生的评分病变的平均增强减少最小（-0.50±0.55）（设置A：-1.10±0.98；设置B：-0.91±0.67，均P < 0.001）。设置A - C中病变符合度的平均评分分别为1.75±1.07、2.19±1.04和2.48±0.91，所有设置之间存在显著差异（均P < 0.001）。
所提出的造影剂信号提取方法在合成增强后图像方面显示出显著改进。在此剂量下，仍有相当比例的图像与参考图像的互换性不足。
