通过分析来自乳头状甲状腺癌（PTC）的双能量计算机断层扫描（DECT）参数，预测中央颈部潜在淋巴结转移（OLNM）的概率。回顾性地收集了经动脉和静脉增强DECT检查并进行中央颈淋巴结清扫（CLND）的病理学上证实为PTC的患者的数据。测量了三个临床特征、三个与形状相关的特征和二十六个DECT衍生参数。应用单变量和多变量分析来选择相关参数并制定正态分布图。总共包括了140个经术前评估阴性诊断为颈部中央淋巴结转移的病例，其中88例为转移（OLNM +），52例为无转移（OLNM -），通过病理学最终确认。 (1)从PTC病灶中得出的前后/横向直径比（A/T）在OLNM + 组和OLNM - 组之间有显著差异（p < 0.05）。 (2)在动脉期，PTC病灶中的碘浓度（ICarterial）、归一化碘浓度（NICarterial）、有效原子数（Zeff-arterial）、电子密度（EDarterial）和能量曲线斜率（karterial）在两个组之间有显著差异（p < 0.05）。在静脉期，只有40 keV下的CT值（HU40keVvenous）有差异（p < 0.05）。 (3)通过建立正态分布图来预测OLNM的概率，并且在训练集和测试集中，AUC、敏感性和特异性分别为0.830、75.0%、76.9%和0.829、65.9%、84.6%。在中央颈部，PTC中的DECT参数结合与形状相关的特征可能可用作OLNM的预测因子。术前影像学评估结合与形状相关的特征和双能量CT参数可以作为甲状腺癌术前手术计划中鉴别中央颈部隐匿淋巴结转移的参考依据。•乳头状甲状腺癌（PTC）患者可能在中央颈部出现隐匿淋巴结转移（OLNM），这在术前影像检查中很难发现。 •双能量CT定量评估比常规CT具有更高的准确性，并可预测PTC中央颈部的OLNM。 •双能量CT定量参数和PTC形态学可作为预测中央颈部OLNM的有用工具，并为个体化治疗提供指导。 © 2023。作者（S），在欧洲放射学社独家许可下。

To predict the probability of occult lymph node metastasis (OLNM) in the central cervical by analyzing the dual-energy computed tomography (DECT) parameters derived from papillary thyroid carcinoma (PTC).Data were retrospectively collected from patients with pathologically confirmed PTC who underwent arterial and venous phases of enhanced DECT with concurrent central neck lymph node dissection (CLND). Three clinical features, three shape-related features, and twenty-six DECT-derived parameters were measured. The univariate and multivariate analyses were applied to select the relevant parameters and develop the nomogram.A total 140 cases with negative diagnosis of cervical central lymph node metastases by preoperative evaluation were included, among which 88 patients with metastasis (OLNM +) and 52 patients without metastasis (OLNM -) were finally confirmed by pathology. (1) Anteroposterior/transverse diameter ratio (A/T) derived from the PTC focus had significant difference between the OLNM + and OLNM - groups (p < 0.05). (2) In the arterial phase, iodine concentration (ICarterial), normalized iodine concentration (NICarterial), effective atomic number (Zeff-arterial), electron density (EDarterial), and slope of energy curve (karterial) from PTC focus showed significant difference (all p < 0.05) between the two groups. In the venous phase, only the CT value under the 40 keV (HU40keVvenous) had differences (p < 0.05). (3) The nomogram was produced to predict the probability of OLNM, and the AUC, sensitivity, and specificity in the training and test cohort were 0.830, 75.0%, 76.9%, and 0.829, 65.9%, 84.6%, respectively.DECT parameters combined with shape-related feature derived from PTC might be used as predictors of OLNM in the central neck.Preoperative imaging evaluation combining shape-related features and dual-energy CT parameters could serve as a reference to discern occult lymph node metastasis in central neck during the surgically planning of papillary thyroid carcinoma.• Papillary thyroid carcinoma (PTC) patients may have occult lymph node metastasis (OLNM) in the central neck, which is extremely difficult to find by preoperative imaging examination. • Dual-energy CT quantitative evaluation has higher accuracy than conventional CT and can predicting OLNM in the central neck of PTC. • Dual-energy CT quantitative parameters and morphology of PTC can serve as a useful tool in predicting OLNM in the central neck, and as a guide for personalized treatment.© 2023. The Author(s), under exclusive licence to European Society of Radiology.