Department of Neurosurgery, University of Tokyo, Tokyo.
Department of Radiology, University of Tokyo, Tokyo.
Int J Radiat Oncol Biol Phys. 2019 Nov 15;105(4):784-791. doi: 10.1016/j.ijrobp.2019.07.011. Epub 2019 Jul 22.
A noninvasive diagnostic method to predict the degree of malignancy accurately would be of great help in glioma management. This study aimed to create a highly accurate machine learning model to perform glioma grading.
Preoperative magnetic resonance imaging acquired for cases of glioma operated on at our institution from October 2014 through January 2018 were obtained retrospectively. Six types of magnetic resonance imaging sequences (T-weighted image, diffusion-weighted image, apparent diffusion coefficient [ADC], fractional anisotropy, and mean kurtosis [MK]) were chosen for analysis; 476 features were extracted semiautomatically for each sequence (2856 features in total). Recursive feature elimination was used to select significant features for a machine learning model that distinguishes glioblastoma from lower-grade glioma (grades 2 and 3).
Fifty-five data sets from 54 cases were obtained (14 grade 2 gliomas, 12 grade 3 gliomas, and 29 glioblastomas), of which 44 and 11 data sets were used for machine learning and independent testing, respectively. We detected 504 features with significant differences (false discovery rate <0.05) between glioblastoma and lower-grade glioma. The most accurate machine learning model was created using 6 features extracted from the ADC and MK images. In the logistic regression, the area under the curve was 0.90 ± 0.05, and the accuracy of the test data set was 0.91 (10 out of 11); using a support vector machine, they were 0.93 ± 0.03 and 0.91 (10 out of 11), respectively (kernel, radial basis function; c = 1.0).
Our machine learning model accurately predicted glioma tumor grade. The ADC and MK sequences produced particularly useful features.
一种能够准确预测恶性程度的非侵入性诊断方法将极大地有助于胶质瘤的管理。本研究旨在创建一个高度准确的机器学习模型来进行胶质瘤分级。
回顾性获取了 2014 年 10 月至 2018 年 1 月期间在我院接受手术治疗的胶质瘤患者的术前磁共振成像数据。选择了 6 种磁共振成像序列（T1 加权像、弥散加权像、表观弥散系数[ADC]、各向异性分数和平均峰度[MK]）进行分析；为每个序列半自动提取 476 个特征（总共 2856 个特征）。递归特征消除用于选择用于区分胶质母细胞瘤和低级别胶质瘤（2 级和 3 级）的机器学习模型的显著特征。
共获得 54 例 55 个数据集（14 例 2 级胶质瘤、12 例 3 级胶质瘤和 29 例胶质母细胞瘤），其中 44 个和 11 个数据集分别用于机器学习和独立测试。我们检测到胶质母细胞瘤与低级别胶质瘤之间有 504 个具有显著差异的特征（错误发现率<0.05）。使用从 ADC 和 MK 图像中提取的 6 个特征创建的最准确的机器学习模型。在逻辑回归中，曲线下面积为 0.90±0.05，测试数据集的准确率为 0.91（11 个中有 10 个）；使用支持向量机时，分别为 0.93±0.03 和 0.91（11 个中有 10 个）（核函数，径向基函数；c=1.0）。
我们的机器学习模型能够准确预测胶质瘤肿瘤分级。ADC 和 MK 序列产生了特别有用的特征。
