基于 MRI 的放射组学模型和列线图预测肝癌患者局部区域治疗结局。
MRI-based radiomics model and nomogram for predicting the outcome of locoregional treatment in patients with hepatocellular carcinoma.
机构信息
Department of Radiology, Beijing Friendship Hospital, Capital Medical University, Beijing, 100050, China.
Department of Radiology, Affiliated Hospital of Changzhi Institute of Traditional Chinese Medicine, Changzhi, 046099, China.
出版信息
BMC Med Imaging. 2023 May 30;23(1):67. doi: 10.1186/s12880-023-01030-5.
BACKGROUND
Prediction of locoregional treatment response is important for further therapeutic strategy in patients with hepatocellular carcinoma. This study aimed to investigate the role of MRI-based radiomics and nomogram for predicting the outcome of locoregional treatment in patients with hepatocellular carcinoma.
METHODS
The initial postoperative MRI after locoregional treatment in 100 patients with hepatocellular carcinoma was retrospectively analysed. The outcome was evaluated according to mRECIST at 6 months. We delineated the tumour volume of interest on arterial phase, portal venous phase and T2WI. The radiomics features were selected by using the independent sample t test or nonparametric Mann‒Whitney U test and the least absolute shrinkage and selection operator. The clinical variables were selected by using univariate analysis and multivariate analysis. The radiomics model and combined model were constructed via multivariate logistic regression analysis. A nomogram was constructed that incorporated the Rad score and selected clinical variables.
RESULTS
Fifty patients had an objective response, and fifty patients had a nonresponse. Nine radiomics features in the arterial phase were selected, but none of the portal venous phase or T2WI radiomics features were predictive of the treatment response. The best radiomics model showed an AUC of 0.833. Two clinical variables (hCRP and therapy method) were selected. The AUC of the combined model was 0.867. There was no significant difference in the AUC between the combined model and the best radiomics model (P = 0.573). Decision curve analysis demonstrated the nomogram has satisfactory predictive value.
CONCLUSIONS
MRI-based radiomics analysis may serve as a promising and noninvasive tool to predict outcome of locoregional treatment in HCC patients, which will facilitate the individualized follow-up and further therapeutic strategies guidance.
背景
预测局部区域治疗反应对于肝细胞癌患者的进一步治疗策略很重要。本研究旨在探讨基于 MRI 的放射组学和列线图在预测肝细胞癌局部区域治疗结果中的作用。
方法
回顾性分析 100 例接受局部区域治疗后初始术后 MRI。根据 6 个月时 mRECIST 评估结果。我们在动脉期、门静脉期和 T2WI 上勾画肿瘤感兴趣区。通过独立样本 t 检验或非参数 Mann-Whitney U 检验和最小绝对收缩和选择算子选择放射组学特征。通过单因素分析和多因素分析选择临床变量。通过多因素逻辑回归分析构建放射组学模型和联合模型。构建一个包含 Rad 评分和选定临床变量的列线图。
结果
50 例患者有客观反应，50 例患者无反应。在动脉期选择了 9 个放射组学特征，但门静脉期或 T2WI 没有放射组学特征可以预测治疗反应。最佳放射组学模型的 AUC 为 0.833。选择了两个临床变量（hCRP 和治疗方法）。联合模型的 AUC 为 0.867。联合模型和最佳放射组学模型的 AUC 之间无显著差异（P=0.573）。决策曲线分析表明，列线图具有令人满意的预测价值。
结论
基于 MRI 的放射组学分析可能是一种很有前途的非侵入性工具，可预测 HCC 患者局部区域治疗的结果，有助于个体化随访和进一步治疗策略的指导。
Zotero 插件