Abstract

Purpose

To quantify how often the LI-RADS v2018 category changed when utilizing major features only, when utilizing major and ancillary features, and when utilizing major and ancillary features excluding gadoxetate-specific ancillary features.

Methods

Retrospective analysis of 100 patients age 18 and older at high risk for hepatocellular carcinoma who had an MRI abdomen performed with intravenous contrast gadoxetate between 1/1/2017 and 3/23/2018. Each examination was reviewed by a body fellowship-trained radiologist. LI-RADS category was assigned to the liver observation after review of major features only. Ancillary features were then reviewed and LI-RADS category assigned both including and excluding ancillary features specific to gadoxetate.

Results

Utilizing all MRI ancillary features, including those specific to gadoxetate, changed the final LI-RADS category in 56.4% of liver observations, the majority an increase or decrease from LR-3. When not including the ancillary features specific to gadoxetate, the final LI-RADS category changed in 30.9% of observations, the majority increasing from LR-3 to LR-4.

Conclusion

Utilizing LI-RADS v2018 ancillary features can significantly alter the final LI-RADS category, especially when using gadoxetate-specific ancillary features. Understanding the correct application of ancillary features for the final LI-RADS category helps implement a more consistent category assessment amongst users.

中文翻译：

---

LI-RADS v2018：利用钆塞酸增强 MRI 的辅助特征来修改最终的 LI-RADS 类别。

摘要

目的

量化 LI-RADS v2018 类别在仅使用主要特征、使用主要和辅助特征以及使用主要和辅助特征（不包括钆塞酸盐特定辅助特征）时的变化频率。

方法

对 100 名 18 岁及以上肝细胞癌高危患者的回顾性分析，这些患者在 2017 年 1 月 1 日至 2018 年 3 月 23 日期间接受了静脉造影剂钆塞酸盐的 MRI 腹部检查。每次检查都由接受过身体奖学金培训的放射科医师进行审查。LI-RADS 类别仅在审查主要特征后分配给肝脏观察。然后审查辅助特征，并分配 LI-RADS 类别，包括和排除特定于钆塞酸盐的辅助特征。

结果

利用所有 MRI 辅助特征，包括特定于钆塞酸盐的特征，在 56.4% 的肝脏观察中改变了最终的 LI-RADS 类别，其中大多数与 LR-3 相比有所增加或减少。当不包括特定于钆塞酸盐的辅助特征时，最终的 LI-RADS 类别在 30.9% 的观察中发生了变化，大多数从 LR-3 增加到 LR-4。

结论

利用 LI-RADS v2018 辅助特征可以显着改变最终的 LI-RADS 类别，尤其是在使用钆塞酸盐特定辅助特征时。了解最终 LI-RADS 类别的辅助特征的正确应用有助于在用户之间实施更一致的类别评估。