Automatic discrimination of different sequences and phases of liver MRI using a dense feature fusion neural network: a preliminary study
Automatic discrimination of different sequences and phases of liver MRI using a dense feature fusion neural network: a preliminary study
复制标题
使用密集特征融合神经网络自动区分肝脏 MRI 的不同序列和阶段：初步研究
DOI：
10.1007/s00261-021-03142-4
复制
发表时间：
2021-05
影响因子：
2.4
通讯作者：
Zheng‑Han Yang
中科院分区：
文献类型：
--
作者：
Shu‑Hui Wang;Jing Du;Hui Xu;Dawei Yang;Yuxiang Ye;Yinan Chen;Yajing Zhu;Te Ba;Chunwang Yuan;Zheng‑Han Yang
目的建立并验证一种能够自动识别肝脏磁共振成像不同序列和不同时相的密集特征融合神经网络。材料与方法采用384例肝脏磁共振检查中的3869个序列和时相，分为训练/验证(n= 2886序列，来自287例)和测试(n= 983序列，来自97例患者)。包括10个非增强序列和增强时相。人工序列识别由两位放射科医生(20年和10年经验)在共识读数中进行，作为参考标准。计算灵敏度、特异度、准确度和受试者工作特征曲线(AUC)下的面积，以评估DFuNN在相同的不可见测试集上的性能。最后对影响模型精度的因素进行了评估。结果融合块提高了DFuNN的性能。采用融合块的DFuNN对测试集中的完整和不完整序列和短语都取得了良好的识别效果。对完整序列和相位输入的平均识别灵敏度为88.06~100%，平均特异度为99.12~99.94%，中位准确率为98.02~99.95%。非肝硬化组的DFuNN预测准确率显著高于肝硬化组(P= 0.0153)。结论DFuNN可以自动、准确地识别特定的平扫序列和增强MRI时相。
PurposeTo develop and validate a dense feature fusion neural network (DFuNN) to automatically recognize different sequences and phases of liver magnetic resonance imaging (MRI).Materials and methodsIn total, 3869 sequences and phases from 384 liver MRI examinations, divided into training/validation (n= 2886 sequences from 287 patients) and test (n= 983 sequences from 97 patients) sets, were used in this retrospective study. Ten unenhanced sequences and enhanced phases were included. Manual sequence recognition, performed by two radiologists (20 and 10 years of experience) in a consensus reading, was used as the reference standard. The sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve (AUC) were calculated to evaluate the performance of the DFuNN on an identical unseen test set. Finally, we evaluated the factors impacting the model precision.ResultsA fusion block improved the performance of the DFuNN. DFuNN with a fusion block achieved good recognition performance for both complete and incomplete sequences and phases in the test set. The average sensitivity of recognition performance for complete sequence and phase inputs ranged from 88.06 to 100%, the average specificity ranged from 99.12 to 99.94%, and the median accuracy ranged from 98.02 to 99.95%. The DFuNN prediction accuracy for patients without cirrhosis were significantly higher than those for patients with cirrhosis (P= 0.0153). No significant difference was found in the accuracy across other factors.ConclusionDFuNN can automatically and accurately identify specific unenhanced MRI sequences and enhanced MRI phases.
登录
查看更多内容
用于心脏 MR 图像左心室容积估计的多视图融合 CNN
影响因子：
4.6
作者：
Luo, Gongning;Dong, Suyu;Zhang, Henggui
通讯作者：
Zhang, Henggui
影响因子：
4.4
作者：
Jumin Zhao;Chen Zhang;Deng-ao Li;Jing Niu
通讯作者：
Jumin Zhao;Chen Zhang;Deng-ao Li;Jing Niu
影响因子：
19.7
作者：
Yasaka, Koichiro;Akai, Hiroyuki;Kiryu, Shigeru
通讯作者：
Kiryu, Shigeru
影响因子：
5.5
作者：
Choi, Jin-Young;Kim, Myeong-Jin;van Cauteren, Marc
通讯作者：
van Cauteren, Marc
影响因子：
3.3
作者：
Kanda, Tomonori;Yoshikawa, Takeshi;Sugimura, Kazuro
通讯作者：
Sugimura, Kazuro