Department of Radiology, Gyeongsang National University School of Medicine and Gyeongsang National University Changwon Hospital, 11 Samjeongja-ro, Seongsan-gu, Changwon, 51472, South Korea.
Department of Surgery, Gyeongsang National University School of Medicine and Gyeongsang National University Changwon Hospital, 11 Samjeongja-ro, Seongsan-gu, Changwon, 51472, South Korea.
Eur Radiol. 2020 May;30(5):2782-2790. doi: 10.1007/s00330-019-06615-9. Epub 2020 Feb 5.
To evaluate the predictive value of virtual monoenergetic images (VMIs) by assessing tumor conspicuity on dual-layer spectral detector CT (SDCT) and correlate tumor conspicuity on VMI with prognostic biomarkers in patients with breast cancer.
Sixty-four patients underwent arterial phase and 90-s delayed phase dual-layer SDCT. A retrospective tumor conspicuity analysis of 14 benign tumors and 65 breast cancers was performed using conventional images (CIs) and VMI at 40 keV (VMI40) on arterial and delayed phase scans (CI, VMI40, CI, VMI40). Mean Hounsfield units (HU) of tumors were measured on VMI40 and VMI40. A receiver operating characteristic (ROC) curve analysis was performed to compare diagnostic accuracy between image sets. Estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER2), and Ki67 levels were evaluated using histopathology. Correlations between VMI analyses and histological characteristics of cancers were analyzed.
Cancers on VMI40 had a significantly higher conspicuity score and mean HU than benign tumors (p < 0.001). VMI40 showed the highest conspicuity for cancers (mean, 3.79) and the greatest area under the ROC curve (0.817; 95% confidence interval 0.745-0.889). VMI40 yielded significantly higher mean HU for cancers than VMI40 (p < 0.001). The conspicuity score and mean HU on VMI40 were significantly higher in cancers with ER negativity, PR negativity, and Ki67 positivity (p < 0.05).
VMI40 may be useful in the diagnosis of breast cancers due to higher tumor conspicuity and better enhancement than VMI40. VMI40 may be beneficial for the prediction of poor breast cancer prognoses.
• VMI40 improved conspicuity of breast cancer than CI. • VMI40yielded higher diagnostic performance of breast cancer than VMI40. • VMI40has an additional benefit in terms of prognosis prediction in patients with breast cancers.
通过评估双层光谱探测器 CT（SDCT）上虚拟单能量图像（VMI）的肿瘤显影程度，评估其在乳腺癌患者中的预测价值，并将 VMI 上的肿瘤显影程度与预后生物标志物相关联。
64 例患者行动脉期和 90 秒延迟期双层 SDCT 检查。对 14 例良性肿瘤和 65 例乳腺癌患者的常规图像（CI）和动脉期及延迟期扫描的 40keV VMI（VMI40）进行回顾性肿瘤显影分析（CI、VMI40、CI、VMI40）。在 VMI40 和 VMI40 上测量肿瘤的平均 CT 值（HU）。通过绘制受试者工作特征（ROC）曲线分析比较图像组之间的诊断准确性。使用组织病理学评估雌激素受体（ER）、孕激素受体（PR）、人表皮生长因子受体 2（HER2）和 Ki67 水平。分析 VMI 分析与癌症组织学特征之间的相关性。
VMI40 上的癌症显影评分和平均 HU 明显高于良性肿瘤（p<0.001）。VMI40 对癌症的显影效果最好（平均值 3.79），ROC 曲线下面积最大（0.817；95%置信区间 0.745-0.889）。VMI40 上癌症的平均 HU 明显高于 VMI40（p<0.001）。ER 阴性、PR 阴性和 Ki67 阳性的癌症 VMI40 的显影评分和平均 HU 显著更高（p<0.05）。
由于 VMI40 比 VMI40 具有更高的肿瘤显影程度和更好的增强效果，因此 VMI40 可能有助于乳腺癌的诊断。VMI40 可能有助于预测乳腺癌不良预后。
VMI40 提高了乳腺癌的显影程度，优于 CI。
VMI40 比 VMI40 具有更高的乳腺癌诊断性能。
VMI40 在预测乳腺癌患者的预后方面具有额外的益处。
