Department of Radiology, 64076Suleyman Demirel University Faculty of Medicine, Suleyman Demirel University Faculty of Medicine, Isparta, Turkey.
Department of Gynecology and Obstetrics, Suleyman Demirel University Faculty of Medicine, Isparta, Turkey.
Acta Radiol. 2022 Aug;63(8):1126-1133. doi: 10.1177/02841851211025853. Epub 2021 Jun 29.
New methods to reduce subjectivity in preoperative magnetic resonance imaging (MRI) staging of endometrial cancer are needed.
To investigate the role of MRI quantitative assessment in staging and risk stratification of endometrial cancer.
Preoperative T2-weighted (T2W) images and diffusion-weighted imaging of 42 patients were analyzed retrospectively by two radiologists. Tumor area ratio (TAR) and tumor volume ratio (TVRseg) were calculated by semi-automatic segmentation of the tumor and uterus on T2W imaging and apparent diffusion coefficient (ADC). TVR was also calculated by the 3D metric method (TVRmetric). Mean ADC was calculated. The patients were allocated to risk groups regarding the stage, grade, and lymphovascular invasion (LVI) status.
TAR, TVRmetric, T2W TVRseg, and ADC TVRseg showed a significant difference between the superficial and deep myometrial invasion groups ( < 0.001). All of these parameters showed a good diagnostic performance for detecting deep myometrial invasion (AUC>0.82), the highest accuracy rate (85%) was found with T2W TVRseg. LVI was significantly associated with TAR ( = 0.002) and T2W TVRseg ( = 0.014), while the cervical invasion was associated with TAR ( = 0.03). ADC was significantly lower in high-grade tumors ( = 0.002). There was a significant difference in ADC ( = 0.002), TAR ( = 0.004), and T2W TVRseg ( = 0.038) between the low- and high-risk groups. AUC of TAR and T2W TVRseg for detecting high-risk groups were 0.80 and 0.77, respectively, while AUC of ADC for the low-risk group was 0.75.
MRI quantitative assessments such as TAR, TVR, and ADC may improve the accuracy of preoperative staging and can help in risk stratification of endometrial cancer.
需要新的方法来减少子宫内膜癌术前磁共振成像(MRI)分期的主观性。
探讨 MRI 定量评估在子宫内膜癌分期和危险分层中的作用。
回顾性分析了 42 例患者的术前 T2 加权(T2W)图像和弥散加权成像。通过对 T2W 图像上肿瘤和子宫的半自动分割,计算肿瘤面积比(TAR)和肿瘤体积比(TVRseg)以及表观扩散系数(ADC)。还通过 3D 度量法(TVRmetric)计算 TVR。根据分期、分级和脉管内侵犯(LVI)状态将患者分配到危险组。
TAR、TVRmetric、T2W TVRseg 和 ADC TVRseg 在浅肌层浸润组和深肌层浸润组之间有显著差异(<0.001)。所有这些参数对检测深肌层浸润均具有良好的诊断性能(AUC>0.82),T2W TVRseg 的准确率最高(85%)。LVI 与 TAR(=0.002)和 T2W TVRseg(=0.014)显著相关,而宫颈浸润与 TAR 相关(=0.03)。高级别肿瘤的 ADC 显著降低(=0.002)。高级别肿瘤组和低级别肿瘤组在 ADC(=0.002)、TAR(=0.004)和 T2W TVRseg(=0.038)方面有显著差异。TAR 和 T2W TVRseg 对高危组的 AUC 分别为 0.80 和 0.77,而 ADC 对低危组的 AUC 为 0.75。
TAR、TVR 和 ADC 等 MRI 定量评估可能提高术前分期的准确性,并有助于子宫内膜癌的危险分层。