功能磁共振成像鉴别肾细胞癌病理类型的初步研究
Functional magnetic resonance imaging for distinguishing type of papillary renal cell carcinoma: a preliminary study.
机构信息
Department of Medical Imaging, Clinical Medical College, Yangzhou University, Yangzhou, China.
出版信息
Br J Radiol. 2021 Oct 1;94(1126):20201315. doi: 10.1259/bjr.20201315. Epub 2021 Sep 7.
OBJECTIVE
To investigate the feasibility of magnetic resonance diffusion kurtosis imaging (DKI) and intravoxel incoherent motion (IVIM) for distinguishing Type 1 and 2 of papillary renal cell carcinoma (PRCC).
METHODS
A total of Type 1 ( = 20) and Type 2 ( = 16) of PRCC were examined by pathology. For DKI and IVIM, mean diffusivity (MD), fractional anisotropy (FA), mean kurtosis (MK), kurtosis anisotropy (KA), radial kurtosis (RK), diffusivity (D), pseudodiffusivity (D*) and perfusion fraction (f) were performed in assessment of type of PRCC.
RESULTS
The mean SNRs of IVIM and DKI images at = 1500 and 2000 s/mm were 8.6 ± 0.8 and 7.8 ± 0.6. Statistically significant differences were observed in MD and D values (1.11 ± 0.23 0.73 ± 0.13, 0.91 ± 0.24 0.49 ± 0.13, < 0.05) between Type 1 and Type 2 of PRCC, while comparable FA, RK, D* and f values were found between Type 1 and Type 2 of PRCC ( > 0.05). Statistically significant differences were observed in MK and KA values (1.23 ± 0.16 1.91 ± 0.26, 1.49 ± 0.19 2.36 ± 0.39, < 0.05) between Type 1 and Type 2 of PRCC. Areas of MD, MK, KA and D values under ROC curves for differentiating Type 1 and Type 2 of PRCC were 0.836, 0.818, 0.881 and 0.766, respectively. Using MD, MK, KA and D values of 0.93, 1.64, 1.94, 0.68 as the threshold value for differentiating Type 1 from Type 2 of PRCC, the best result obtained had a sensitivity of 85.0%, 80.0%, 90.0%, 85.0%, a specificity 75.0%, 68.7%, 87.5%, 81.2%, and an accuracy of 83.3%, 80.5%, 88.9%, 86.1%, respectively.
CONCLUSION
DKI and IVIM are feasible techniques for distinguishing type of PRCC, given an adequate SNR of IVIM and DKI images.
ADVANCES IN KNOWLEDGE
- MD and D values are higher for Type 1 of PRCC and lower for Type 2 of PRCC.2. MK and KA values are higher for Type 2 of PRCC and lower for Type 1 of PRCC.3. DKI and IVIM can be used as clinical biomarker for PRCC type's differential diagnosis, given an adequate SNR.
目的
探讨磁共振扩散峰度成像（DKI）和体素内不相干运动（IVIM）在鉴别 1 型和 2 型乳头状肾细胞癌（PRCC）中的可行性。
方法
对经病理证实的 20 例 1 型和 16 例 2 型 PRCC 进行检查。对于 DKI 和 IVIM，在评估 PRCC 类型时，分别测量平均扩散系数（MD）、各向异性分数（FA）、平均峰度（MK）、峰度各向异性（KA）、径向峰度（RK）、扩散系数（D）、假性扩散系数（D*）和灌注分数（f）。
结果
1500 和 2000 s/mm 时 IVIM 和 DKI 图像的平均信噪比（SNR）分别为 8.6 ± 0.8 和 7.8 ± 0.6。1 型和 2 型 PRCC 之间 MD 和 D 值存在显著差异（1.11 ± 0.23 0.73 ± 0.13，0.91 ± 0.24 0.49 ± 0.13， < 0.05），而 1 型和 2 型 PRCC 之间 FA、RK、D*和 f 值相当（ > 0.05）。1 型和 2 型 PRCC 之间 MK 和 KA 值存在显著差异（1.23 ± 0.16 1.91 ± 0.26，1.49 ± 0.19 2.36 ± 0.39， < 0.05）。MD、MK、KA 和 D 值的 ROC 曲线下面积分别为 0.836、0.818、0.881 和 0.766，用于鉴别 1 型和 2 型 PRCC。当 MD、MK、KA 和 D 值的截断值分别为 0.93、1.64、1.94 和 0.68 时，1 型和 2 型 PRCC 的最佳鉴别结果具有 85.0%、80.0%、90.0%和 85.0%的灵敏度，75.0%、68.7%、87.5%和 81.2%的特异性，83.3%、80.5%、88.9%和 86.1%的准确性。
结论
在 IVIM 和 DKI 图像具有足够 SNR 的情况下，DKI 和 IVIM 是鉴别 PRCC 类型的可行技术。
知识的进步
- 1 型 PRCC 的 MD 和 D 值较高，2 型 PRCC 的 MD 和 D 值较低。2. 2 型 PRCC 的 MK 和 KA 值较高，1 型 PRCC 的 MK 和 KA 值较低。3. 在具有足够 SNR 的情况下，DKI 和 IVIM 可用作 PRCC 类型鉴别诊断的临床生物标志物。
Zotero 插件