Department of Radiology, Changhai Hospital, The Navy Military Medical University, 168 Changhai Road, Shanghai, 200433, China.
Department of Pathology, Changhai Hospital, The Navy Military Medical University, Shanghai, China.
Abdom Radiol (NY). 2021 Feb;46(2):667-680. doi: 10.1007/s00261-020-02706-0. Epub 2020 Aug 17.
We aimed to explore the relationship between the magnetic resonance imaging (MRI) radiomic score (rad-score) and the grades of non-functioning pancreatic neuroendocrine tumors (NF-pNETs) and evaluate the potential of the calculated MRI rad-score to differentiate grade 1 from grade 2/3 NF-pNETs.
This retrospective study assessed 157 patients with surgically resected, pathologically confirmed NF-pNETs who underwent magnetic resonance scans from November 2012 to December 2019. Radiomic features were extracted from arterial and portal venous MRI. The least absolute shrinkage and selection operator method were used to select the features. Multivariate logistic regression models were used to analyze the association between the MRI rad-score and NF-pNET grades. The MRI rad-score performance was assessed based on its discriminative ability and clinical usefulness.
The MRI rad-score, which consisted of seven selected features, was significantly associated with the NF-pNET grades. Every 1-point increase in the rad-score was associated with a 35% increased risk of grade 2/3 disease. The score also showed high accuracy (area under the curve = 0.775). The best cut-off point for maximal sensitivity and specificity was at 0.41. In the decision curves, when the threshold probability was higher than 0.3, the rad-score used in this study to distinguish grades 1 and 2/3 NF-pNETs offered more benefits than the use of a treat-all-patients or a treat-none scheme.
The MRI rad-score showed a significant association with the grades of NF-pNETs. Thus, it may be used as a valuable non-invasive tool for differential NF-pNET grading.
本研究旨在探讨磁共振成像（MRI）放射组学评分（rad-score）与无功能性胰腺神经内分泌肿瘤（NF-pNETs）分级的关系，并评估计算出的 MRI rad-score 区分 1 级和 2/3 级 NF-pNETs 的潜力。
本回顾性研究纳入了 2012 年 11 月至 2019 年 12 月期间接受 MRI 扫描并经手术切除和病理证实为 NF-pNETs 的 157 例患者。从动脉期和门静脉期 MRI 中提取放射组学特征。使用最小绝对收缩和选择算子法选择特征。使用多变量逻辑回归模型分析 MRI rad-score 与 NF-pNET 分级之间的关系。根据判别能力和临床实用性评估 MRI rad-score 的性能。
由 7 个选定特征组成的 MRI rad-score 与 NF-pNET 分级显著相关。rad-score 每增加 1 分，2/3 级疾病的风险就会增加 35%。该评分也具有较高的准确性（曲线下面积=0.775）。最大灵敏度和特异性的最佳截断点为 0.41。在决策曲线中，当阈值概率高于 0.3 时，本研究中使用的 rad-score 用于区分 1 级和 2/3 级 NF-pNETs 的效果优于使用全部治疗或不治疗方案。
MRI rad-score 与 NF-pNET 的分级显著相关。因此，它可能成为一种有价值的非侵入性工具，用于 NF-pNET 的分级鉴别。
