Machine Learning methods for Quantitative Radiomic Biomarkers.
Machine Learning methods for Quantitative Radiomic Biomarkers.
复制标题
DOI：
10.1038/srep13087
复制
发表时间：
2015-08-17
影响因子：
4.6
通讯作者：
Aerts HJWL
中科院分区：
文献类型：
--
作者：
Parmar C;Grossmann P;Bussink J;Lambin P;Aerts HJWL
放射组学提取和挖掘大量的医学影像特征，量化肿瘤的表型特征。高度精确和可靠的机器学习方法可以推动放射学在临床护理中的应用取得成功。在这项放射组学研究中，研究了14种特征选择方法和12种分类方法在预测总体生存方面的性能和稳定性。从464例肺癌患者的术前CT图像中提取440个放射学特征。为了确保对不同机器学习方法的公正评估，使用了公开可用的实现以及报告的参数配置。此外，我们使用了两个独立的放射学队列进行训练（n = 310例患者）和验证（n = 154例患者）。我们发现基于Wilcoxon检验的特征选择方法WLCX（稳定性= 0.84±0.05,AUC = 0.65±0.02）和分类方法随机森林RF （RSD = 3.52%, AUC = 0.66±0.03）具有最高的预测性能，对数据扰动具有较高的稳定性。我们的变异性分析表明，分类方法的选择是绩效差异的最主要来源（占总方差的34.21%）。确定放射组学应用的最佳机器学习方法是朝着稳定和临床相关的放射组学生物标志物迈出的关键一步，为临床实践中量化和监测肿瘤表型特征提供了一种非侵入性方法。
Radiomics extracts and mines large number of medical imaging features quantifying tumor phenotypic characteristics. Highly accurate and reliable machine-learning approaches can drive the success of radiomic applications in clinical care. In this radiomic study, fourteen feature selection methods and twelve classification methods were examined in terms of their performance and stability for predicting overall survival. A total of 440 radiomic features were extracted from pre-treatment computed tomography (CT) images of 464 lung cancer patients. To ensure the unbiased evaluation of different machine-learning methods, publicly available implementations along with reported parameter configurations were used. Furthermore, we used two independent radiomic cohorts for training (n = 310 patients) and validation (n = 154 patients). We identified that Wilcoxon test based feature selection method WLCX (stability = 0.84 ± 0.05, AUC = 0.65 ± 0.02) and a classification method random forest RF (RSD = 3.52%, AUC = 0.66 ± 0.03) had highest prognostic performance with high stability against data perturbation. Our variability analysis indicated that the choice of classification method is the most dominant source of performance variation (34.21% of total variance). Identification of optimal machine-learning methods for radiomic applications is a crucial step towards stable and clinically relevant radiomic biomarkers, providing a non-invasive way of quantifying and monitoring tumor-phenotypic characteristics in clinical practice.