机构：[1]Department of Interventional Neuroradiology, Beijing Neurosurgical Institute and Beijing Tiantan Hospital of Capital Medical University, China重点科室医技科室研究所放射科放射科北京市神经外科研究所首都医科大学附属天坛医院[2]Beijing Neurointerventional Engineering Center, China
Background and Purpose- Discrimination of the stability of intracranial aneurysms is critical for determining the treatment strategy, especially in small aneurysms. This study aims to evaluate the feasibility of applying machine learning for predicting aneurysm stability with radiomics-derived morphological features. Methods- Morphological features of 719 aneurysms were extracted from PyRadiomics, of which 420 aneurysms with Maximum3DDiameter ranging from 4 mm to 8 mm were enrolled for analysis. The stability of these aneurysms and other clinical characteristics were reviewed from the medical records. Based on the morphologies with/without clinical features, machine learning models were constructed and compared to define the morphological determinants and screen the optimal model for predicting aneurysm stability. The effect of clinical characteristics on the morphology of unstable aneurysms was analyzed. Results- Twelve morphological features were automatically extracted from PyRadiomics implemented in Python for each aneurysm. Lasso regression defined Flatness as the most important morphological feature to predict aneurysm stability, followed by SphericalDisproportion, Maximum2DDiameterSlice, and SurfaceArea. SurfaceArea (odds ratio [OR], 0.697; 95% CI, 0.476-0.998), SphericalDisproportion (OR, 1.730; 95% CI, 1.143-2.658), Flatness (OR, 0.584; 95% CI, 0.374-0.894), Hyperlipemia (OR, 2.410; 95% CI, 1.029-5.721), Multiplicity (OR, 0.182; 95% CI, 0.082-0.380), Location at middle cerebral artery (OR, 0.359; 95% CI, 0.134-0.902), and internal carotid artery (OR, 0.087; 95% CI, 0.030-0.211) were enrolled into the final prediction model. In terms of performance, the area under curve of the model reached 0.853 (95% CI, 0.767-0.940). For unstable aneurysms, Compactness1 (P=0.035), Compactness2 (P=0.036), Sphericity (P=0.035), and Flatness (P=0.010) were low, whereas SphericalDisproportion (P=0.034) was higher in patients with hypertension. Conclusions- Morphological features extracted from PyRadiomics can be used for aneurysm stratification. Flatness is the most important morphological determinant to predict aneurysm stability. Our model can be used to predict aneurysm stability. Unstable aneurysm is more irregular in patients with hypertension.
基金：
National Key Research and Development Program of China [2017YFB1304400]
第一作者机构：[1]Department of Interventional Neuroradiology, Beijing Neurosurgical Institute and Beijing Tiantan Hospital of Capital Medical University, China[2]Beijing Neurointerventional Engineering Center, China
通讯作者：
通讯机构：[1]Department of Interventional Neuroradiology, Beijing Neurosurgical Institute and Beijing Tiantan Hospital of Capital Medical University, China[2]Beijing Neurointerventional Engineering Center, China[*1]Department of Interventional Neuroradiology, Beijing Neurosurgical Institute and Beijing Tiantan Hospital of Capital Medical University, Beijing, China.
推荐引用方式(GB/T 7714)：
QingLin Liu,Peng Jiang,YuHua Jiang,et al.Prediction of Aneurysm Stability Using a Machine Learning Model Based on PyRadiomics-Derived Morphological Features[J].STROKE.2019,50(9):2314-2321.doi:10.1161/STROKEAHA.119.025777.
APA：
QingLin Liu,Peng Jiang,YuHua Jiang,HuiJian Ge,ShaoLin Li...&YouXiang Li.(2019).Prediction of Aneurysm Stability Using a Machine Learning Model Based on PyRadiomics-Derived Morphological Features.STROKE,50,(9)
MLA：
QingLin Liu,et al."Prediction of Aneurysm Stability Using a Machine Learning Model Based on PyRadiomics-Derived Morphological Features".STROKE 50..9(2019):2314-2321
医学