当前位置: X-MOL 学术Neuroradiology › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Quantitative dynamic contrast-enhanced MRI and readout segmentation of long variable echo-trains diffusion-weighted imaging in differentiating parotid gland tumors
Neuroradiology ( IF 2.8 ) Pub Date : 2021-07-09 , DOI: 10.1007/s00234-021-02758-z
Nan Huang 1 , Zebin Xiao 2 , Yu Chen 1 , Dejun She 1 , Wei Guo 1 , Xiefeng Yang 1 , Qi Chen 1 , Dairong Cao 1 , Tanhui Chen 1
Affiliation  

Purpose

To evaluate the ability of quantitative dynamic contrast-enhanced (DCE)-MRI and readout segmentation of long variable echo-trains diffusion-weighted imaging (RESOLVE-DWI) in differentiating parotid tumors (PTs) with different histological types.

Methods

In this retrospective study, 123 patients with 145 histologically proven PTs who underwent both RESOLVE-DWI and DCE-MRI were enrolled including 51 pleomorphic adenomas (PAs), 52 Warthin’s tumors (WTs), 27 other benign neoplasms (OBNs), and 15 malignant tumors (MTs). Quantitative parameters of DCE-MRI (Ktrans, Kep, and Ve) and the apparent diffusion coefficient (ADC) of lesions were calculated and analyzed. Kruskal–Wallis tests with Dunn-Bonferroni correction, logistic regression analyses, and receiver operating characteristic curve were used for statistical analyses.

Results

PAs exhibited a lowest Ktrans among these four PTs. WTs demonstrated the highest Kep and lowest Ve values. WTs and MTs showed lower ADCmin values than PAs and OBNs. The combination of Kep and Ve provided 98.1% sensitivity, 85% specificity, and 98.7% accuracy for differentiating WTs from the other three PTs. The ADCmin cutoff value of ≤ 0.826 yielded 80.0% sensitivity, 92.3% specificity, and 90.3% accuracy for the differentiation of MTs from PAs and OBNs. Ktrans with a cutoff value of ≤ 0.185 achieved a sensitivity, specificity, and accuracy of 84.3, 70.4, and 79.5%, respectively, for discriminating PAs from OBNs.

Conclusion

The combination of quantitative DCE-MRI and RESOLVE-DWI is beneficial for characterizing four histological types of PTs.



中文翻译:

定量动态对比增强 MRI 和长可变回波序列扩散加权成像的读出分割在区分腮腺肿瘤中

目的

评估定量动态对比增强 (DCE)-MRI 和长可变回波序列弥散加权成像 (RESOLVE-DWI) 的读出分割在区分不同组织学类型的腮腺肿瘤 (PT) 中的能力。

方法

在这项回顾性研究中,纳入了 123 名患者,其中 145 名经组织学证实的 PT 接受了 RESOLVE-DWI 和 DCE-MRI,包括 51 名多形性腺瘤 (PAs)、52 名 Warthin 瘤 (WTs)、27 名其他良性肿瘤 (OBNs) 和 15 名恶性肿瘤(MT)。计算并分析DCE-MRI的定量参数(K trans、K ep和V e)和病灶表观扩散系数(ADC)。使用带有 Dunn-Bonferroni 校正的 Kruskal-Wallis 检验、逻辑回归分析和接受者操作特征曲线进行统计分析。

结果

PA在这四个 PT 中表现出最低的 K trans。WT 表现出最高的 K ep和最低的 V e值。WT 和 MT 的 ADC最小值低于 PA 和 OBN。K ep和 V e的组合为区分 WT 与其他三个 PT 提供了 98.1% 的灵敏度、85% 的特异性和 98.7% 的准确度。≤ 0.826的 ADC最小截止值产生了 80.0% 的灵敏度、92.3% 的特异性和 90.3% 的准确度,用于区分 MT 与 PA 和 OBN。截断值≤ 0.185 的K trans 分别实现了 84.3%、70.4% 和 79.5% 的灵敏度、特异性和准确度,用于区分 PA 和 OBN。

结论

定量 DCE-MRI 和 RESOLVE-DWI 的组合有利于表征四种组织学类型的 PT。

更新日期:2021-07-09
down
wechat
bug