Department of Minimal Invasive Intervention Radiology, Ganzhou People's Hospital, Ganzhou, China.
Department of Cardiology, The First Affiliated Hospital of Jinan university, Guanghzhou, China.
Cancer Med. 2023 Sep;12(17):17529-17540. doi: 10.1002/cam4.6277. Epub 2023 Sep 11.
To develop a deep learning radiomics of multiparametric magnetic resonance imaging (DLRMM)-based model that incorporates preoperative and postoperative signatures for prediction of local tumor progression (LTP) after thermal ablation (TA) in hepatocellular carcinoma (HCC).
From May 2017 to October 2021, 417 eligible patients with HCC were retrospectively enrolled from three hospitals (one primary cohort [PC, n = 189] and two external test cohorts [ETCs][n = 135, 93]). DLRMM features were extracted from T1WI + C, T2WI, and DWI using ResNet18 model. An integrative model incorporating the DLRMM signature with clinicopathologic variables were further built to LTP risk stratification. The performance of these models were compared by areas under receiver operating characteristic curve (AUC) using DeLong test.
A total of 1668 subsequences and 31,536 multiparametric MRI slice including T1WI, T2WI, and DWI were collected simultaneously. The DLRMM signatures were extracted from tumor and ablation zone, respectively. Ablative margin, multiple tumors, and tumor abutting major vessels were regarded as risk factors for LTP in clinical model. The AUC of DLRMM model were 0.864 in PC, 0.843 in ETC1, and 0.858 in ETC2, which was higher significantly than those in clinical model (p < 0.001). After integrating clinical variable, DLRMM model obtained significant improvement with AUC of 0.870-0.869 in three cohorts (all, p < 0.001), which can provide the risk stratification for overall survival of HCC patients.
The DLRMM model is essential to identify LTP risk of HCC patients who underwent TA and may potentially benefit personalized decision-making.
开发一种基于多参数磁共振成像（MP-MRI）深度学习放射组学（DLRMM）的模型，该模型纳入术前和术后特征，以预测肝癌（HCC）热消融（TA）后局部肿瘤进展（LTP）。
本研究回顾性纳入 2017 年 5 月至 2021 年 10 月期间来自三家医院的 417 例 HCC 患者（一个主要队列[PC，n=189]和两个外部测试队列[ETC][n=135，93]）。使用 ResNet18 模型从 T1WI+C、T2WI 和 DWI 中提取 DLRMM 特征。进一步构建整合了 DLRMM 特征和临床病理变量的综合模型，以进行 LTP 风险分层。使用 DeLong 检验比较这些模型的受试者工作特征曲线（ROC）下面积（AUC）。
共采集了 1668 个序列和 31536 个包括 T1WI、T2WI 和 DWI 的多参数 MRI 切片。从肿瘤和消融区分别提取 DLRMM 特征。边缘消融、多发肿瘤和肿瘤毗邻大血管被认为是临床模型中 LTP 的危险因素。在 PC 中，DLRMM 模型的 AUC 为 0.864，在 ETC1 中为 0.843，在 ETC2 中为 0.858，均显著高于临床模型（p<0.001）。在整合临床变量后，DLRMM 模型在三个队列中的 AUC 均显著提高，为 0.870-0.869（均，p<0.001），可提供 HCC 患者总生存风险分层。
DLRMM 模型对于识别接受 TA 的 HCC 患者的 LTP 风险至关重要，并且可能有助于个性化决策。
