Department of Radiology, Peking University Third Hospital, Haidian District, Beijing, People's Republic of China.
Huiying Medical Technology (Beijing) Co., Ltd, Beijing, China.
J Magn Reson Imaging. 2021 Oct;54(4):1303-1311. doi: 10.1002/jmri.27637. Epub 2021 May 12.
Radiomics has shown promising results in the diagnosis, efficacy, and prognostic assessments of multiple myeloma (MM). However, little evidence exists on the utility of radiomics in predicting a high-risk cytogenetic (HRC) status in MM.
To develop and test a magnetic resonance imaging (MRI)-based radiomics model for predicting an HRC status in MM patients.
Retrospective.
Eighty-nine MM patients (HRC [n: 37] and non-HRC [n: 52]).
FIELD STRENGTH/SEQUENCE: A 3.0 T; fast spin-echo (FSE): T1-weighted image (T1WI) and fat-suppression T2WI (FS-T2WI).
Overall, 1409 radiomics features were extracted from each volume of interest drawn by radiologists. Three sequential feature selection steps-variance threshold, SelectKBest, and least absolute shrinkage selection operator-were repeated 10 times with 5-fold cross-validation. Radiomics models were constructed with the top three frequency features of T WI/T WI/two-sequence MRI (T WI and FS-T WI). Radiomics models, clinical data (age and visually assessed MRI pattern), or radiomics combined with clinical data were used with six classifiers to distinguish between HRC and non-HRC statuses. Six classifiers used were support vector machine, random forest, logistic regression (LR), decision tree, k-nearest neighbor, and XGBoost. Model performance was evaluated with area under the curve (AUC) values.
Mann-Whitney U-test, Chi-squared test, Z test, and DeLong method.
The LR classifier performed better than the other classifiers based on different data (AUC: 0.65-0.82; P < 0.05). The two-sequence MRI models performed better than the other data models using different classifiers (AUC: 0.68-0.82; P < 0.05). Thus, the LR two-sequence model yielded the best performance (AUC: 0.82 ± 0.02; sensitivity: 84.1%; specificity: 68.1%; accuracy: 74.7%; P < 0.05).
The LR-based machine learning method appears superior to other classifier methods for assessing HRC in MM. Radiomics features based on two-sequence MRI showed good performance in differentiating HRC and non-HRC statuses in MM.
3 TECHNICAL EFFICACY: Stage 2.
放射组学在多发性骨髓瘤（MM）的诊断、疗效和预后评估方面显示出了良好的效果。然而，关于放射组学在预测 MM 高危细胞遗传学（HRC）状态方面的应用，目前证据有限。
开发并测试一种基于磁共振成像（MRI）的放射组学模型，用于预测 MM 患者的 HRC 状态。
回顾性。
89 例 MM 患者（HRC [n=37]和非-HRC [n=52]）。
磁场强度/序列：3.0 T；快速自旋回波（FSE）：T1 加权图像（T1WI）和脂肪抑制 T2WI（FS-T2WI）。
总体而言，由放射科医生绘制的每个感兴趣容积中提取了 1409 个放射组学特征。通过重复 10 次、采用 5 倍交叉验证的方差阈值、SelectKBest 和最小绝对收缩和选择算子（LASSO）3 个连续的特征选择步骤，选择了 T1WI/T1WI/双序列 MRI（T1WI 和 FS-T2WI）的前三个频率特征。使用支持向量机、随机森林、逻辑回归（LR）、决策树、k-最近邻和 XGBoost 等 6 种分类器，基于 T1WI/T1WI/双序列 MRI（T1WI 和 FS-T2WI）的前三个频率特征构建放射组学模型，以区分 HRC 和非-HRC 状态。使用 6 种分类器分别对 HRC 和非-HRC 状态进行分类，6 种分类器分别为支持向量机、随机森林、逻辑回归、决策树、k-最近邻和 XGBoost。使用曲线下面积（AUC）值评估模型性能。
Mann-Whitney U 检验、卡方检验、Z 检验和 Delong 方法。
基于不同数据，LR 分类器的表现优于其他分类器（AUC：0.65-0.82；P<0.05）。基于不同分类器，双序列 MRI 模型的表现优于其他数据模型（AUC：0.68-0.82；P<0.05）。因此，LR 双序列模型的表现最佳（AUC：0.82±0.02；敏感性：84.1%；特异性：68.1%；准确性：74.7%；P<0.05）。
基于 LR 的机器学习方法似乎优于其他分类器方法，用于评估 MM 中的 HRC。基于双序列 MRI 的放射组学特征在区分 MM 中的 HRC 和非-HRC 状态方面表现良好。
3 级技术功效：2 级。
