Department of Radiology and Research Institute of Radiological Science, College of Medicine, Yonsei University College of Medicine, 50 Yonsei-ro, Seodaemun-gu, Seoul, 120-752, Korea.
Department of Neurosurgery, Yonsei University College of Medicine, Seoul, Korea.
Eur Radiol. 2018 Sep;28(9):3832-3839. doi: 10.1007/s00330-018-5368-4. Epub 2018 Apr 6.
To evaluate the diagnostic performance of magnetic resonance (MR) radiomics-based machine-learning algorithms in differentiating primary central nervous system lymphoma (PCNSL) from non-necrotic atypical glioblastoma (GBM).
Seventy-seven patients (54 individuals with PCNSL and 23 with non-necrotic atypical GBM), diagnosed from January 2009 to April 2017, were enrolled in this retrospective study. A total of 6,366 radiomics features, including shape, volume, first-order, texture, and wavelet-transformed features, were extracted from multi-parametric (post-contrast T1- and T2-weighted, and fluid attenuation inversion recovery images) and multiregional (enhanced and non-enhanced) tumour volumes. These features were subjected to recursive feature elimination and random forest (RF) analysis with nested cross-validation. The diagnostic abilities of a radiomics machine-learning classifier, apparent diffusion coefficient (ADC), and three readers, who independently classified the tumours based on conventional MR sequences, were evaluated using receiver operating characteristic (ROC) analysis. Areas under the ROC curves (AUC) of the radiomics classifier, ADC value, and the radiologists were compared.
The mean AUC of the radiomics classifier was 0.921 (95 % CI 0.825-0.990). The AUCs of the three readers and ADC were 0.707 (95 % CI 0.622-0.793), 0.759 (95 %CI 0.656-0.861), 0.695 (95 % CI 0.590-0.800) and 0.684 (95 % CI0.560-0.809), respectively. The AUC of the radiomics-based classifier was significantly higher than those of the three readers and ADC (p< 0.001 for all).
Large-scale radiomics with a machine-learning algorithm can be useful for differentiating PCNSL from atypical GBM, and yields a better diagnostic performance than human radiologists and ADC values.
• Machine-learning algorithm radiomics can help to differentiate primary central PCNSL from GBM. • This approach yields a higher diagnostic accuracy than visual analysis by radiologists. • Radiomics can strengthen radiologists' diagnostic decisions whenever conventional MRI sequences are available.
评估基于磁共振成像（MR）放射组学的机器学习算法在鉴别原发性中枢神经系统淋巴瘤（PCNSL）与非坏死性非典型胶质母细胞瘤（GBM）中的诊断性能。
本回顾性研究纳入了 2009 年 1 月至 2017 年 4 月间诊断为 PCNSL（54 例）和非坏死性非典型 GBM（23 例）的 77 例患者。从多参数（增强后 T1 和 T2 加权、液体衰减反转恢复图像）和多区域（增强和非增强）肿瘤体积中提取了 6366 个放射组学特征，包括形状、体积、一阶、纹理和小波变换特征。采用嵌套交叉验证对这些特征进行递归特征消除和随机森林（RF）分析。使用受试者工作特征（ROC）曲线分析评估放射组学机器学习分类器、表观扩散系数（ADC）和三位独立基于常规 MR 序列对肿瘤进行分类的阅片者的诊断能力。比较了放射组学分类器、ADC 值和阅片者的 ROC 曲线下面积（AUC）。
放射组学分类器的平均 AUC 为 0.921（95%CI 0.825-0.990）。三位阅片者和 ADC 的 AUC 分别为 0.707（95%CI 0.622-0.793）、0.759（95%CI 0.656-0.861）、0.695（95%CI 0.590-0.800）和 0.684（95%CI0.560-0.809）。放射组学分类器的 AUC 明显高于三位阅片者和 ADC（均 P<0.001）。
基于放射组学的机器学习算法可用于鉴别 PCNSL 与非典型 GBM，其诊断性能优于人类阅片者和 ADC 值。
基于机器学习算法的放射组学有助于鉴别原发性中枢神经系统 PCNSL 与 GBM。
与放射科医生的视觉分析相比，该方法具有更高的诊断准确性。
只要有常规 MRI 序列，放射组学就能增强放射科医生的诊断决策能力。
