Department of Radiology, Changhai Hospital, No.168 Changhai Road, Shanghai, 200433, China.
Huiying Medical Technology Co., Ltd, Beijing, China.
BMC Med Imaging. 2021 Feb 16;21(1):30. doi: 10.1186/s12880-021-00560-0.
To validate and compare various MRI-based radiomics models to evaluate treatment response to neoadjuvant chemoradiotherapy (nCRT) of rectal cancer.
A total of 80 patients with locally advanced rectal cancer (LARC) who underwent surgical resection after nCRT were enrolled retrospectively. Rectal MR images were scanned pre- and post-nCRT. The radiomics features were extracted from T2-weighted images, then reduced separately by least absolute shrinkage and selection operator (LASSO) and principal component analysis (PCA). Four classifiers of Logistic Regression, Random Forest (RF), Decision Tree and K-nearest neighbor (KNN) models were constructed to assess the tumor regression grade (TRG) and pathologic complete response (pCR), respectively. The diagnostic performances of models were determined with leave-one-out cross-validation by generating receiver operating characteristic curves and decision curve analysis.
Three features related to the TRG and 11 features related to the pCR were obtained by LASSO. Top five principal components representing a cumulative contribution of 80% to overall features were selected by PCA. For TRG, the area under the curve (AUC) of RF model was 0.943 for LASSO and 0.930 for PCA, higher than other models (P < 0.05 for both). As for pCR, the AUCs of KNN for LASSO and PCA were 0.945 and 0.712, higher than other models (P < 0.05 for both). The DCA showed that LASSO algorithm was clinically superior to PCA.
MRI-based radiomics models demonstrated good performance for evaluating the treatment response of LARC after nCRT and LASSO algorithm yielded more clinical benefit.
为了验证和比较各种基于 MRI 的放射组学模型，以评估新辅助放化疗（nCRT）治疗直肠癌的治疗反应。
回顾性纳入 80 例接受 nCRT 后行手术切除的局部进展期直肠癌（LARC）患者。对 nCRT 前后的直肠磁共振成像（MRI）进行扫描。从 T2 加权图像中提取放射组学特征，然后分别通过最小绝对值收缩和选择算子（LASSO）和主成分分析（PCA）进行降维。构建逻辑回归、随机森林（RF）、决策树和 K 最近邻（KNN）模型等四种分类器，分别评估肿瘤退缩分级（TRG）和病理完全缓解（pCR）。通过生成接受者操作特征曲线和决策曲线分析，采用留一交叉验证法确定模型的诊断性能。
LASSO 获得了与 TRG 相关的 3 个特征和与 pCR 相关的 11 个特征。PCA 选择了代表总体特征 80%累积贡献的前 5 个主成分。对于 TRG，LASSO 模型的 RF 曲线下面积（AUC）为 0.943，PCA 为 0.930，均高于其他模型（均 P<0.05）。对于 pCR，LASSO 和 PCA 的 KNN 的 AUC 分别为 0.945 和 0.712，均高于其他模型（均 P<0.05）。DCA 显示 LASSO 算法具有更高的临床获益。
基于 MRI 的放射组学模型在评估 nCRT 治疗 LARC 后的治疗反应方面表现良好，LASSO 算法具有更高的临床获益。
