[1]胡小龙,邓朋,唐晓宇,等.基于临床-影像组学特征的机器学习模型预测颅内动脉瘤的破裂风险[J].中国临床神经外科杂志,2023,28(09):549-553.[doi:10.13798/j.issn.1009-153X.2023.09.002]
　HU Xiao-long,DENG Peng,TANG Xiao-yu,et al.Prediction of rupture risk of intracranial aneurysms using a machine learning model based on clinico-radiomic features[J].,2023,28(09):549-553.[doi:10.13798/j.issn.1009-153X.2023.09.002]
点击复制

基于临床-影像组学特征的机器学习模型预测颅内动脉瘤的破裂风险()
分享到：

《中国临床神经外科杂志》[ISSN:1009-153X/CN:42-1603/TN]

卷:
28
期数:
2023年09期
页码:
549-553
栏目:
论著
出版日期:
2023-09-30

文章信息/Info

Title:
Prediction of rupture risk of intracranial aneurysms using a machine learning model based on clinico-radiomic features
文章编号:
1009-153X（2023）09-0549-05
作者:
胡小龙; 邓朋; 唐晓宇; 马冕; 钱锦宏; 吴刚; 成之奇; 龚宇珲; 吴建东; 丁志良
215000江苏苏州，南京医科大学附属苏州医院神经外科（胡小龙、邓朋、唐晓宇、马冕、钱锦宏、吴刚、成之奇、龚宇珲、吴建东、丁志良）
Author(s):
HU Xiao-long;  DENG Peng;  TANG Xiao-yu;  MA Ming;  QIAN Jin-hong;  WU Gang;  CHENG Zhi-qi;  GONG Yu-hui;  WU Jian-dong;  DING Zhi-liang
Department of Neurosurgery, Suzhou Hospital Affiliated to Nanjing Medical University, Suzhou 215000, China
关键词:
颅内动脉瘤; 影像组学; 破裂风险; 机器学习模型
Keywords:
Intracranial aneurysms;  Radiomics;  Machine learning model;  Risk of intracranial aneurysm rupture
分类号:
R 743.9
DOI:
10.13798/j.issn.1009-153X.2023.09.002
文献标志码:
A
摘要:
目的 探讨基于临床-影像组学特征的机器学习模型预测颅内动脉瘤破裂风险的效能。方法 回顾性分析2019年1月至2022年12月收治的153例颅内动脉瘤的临床资料。利用单因素分析筛选出影响动脉瘤破裂的临床特征；收集DICOM格式影像数据，利用3D Slicer软件对载瘤动脉进行三维重建与分割，利用radiomics插件提取动脉瘤的形态学特征，并利用LASSO算法筛选影响动脉瘤破裂的重要形态学特征；构建基于临床和形态学特征的机器学习模型，计算各模型预测颅内动脉瘤破裂风险的曲线下面积、准确度、精准度、灵敏度、特异度。结果 153例中，破裂动脉瘤43例（破裂组），未破裂动脉瘤110例（未破裂组）。与未破裂组相比，破裂组高血压病人占比较高（P<0.05）。与未破裂组相比较，破裂组动脉瘤形态学特征的伸长率、球形度明显缩小（P<0.05），而动脉瘤最小轴径、最大轴径、冠状面最大直径、最大三维直径、网格体积、表面积、表面积体积比和体素体积明显增大（P<0.05）。纳入分析的14个形态学特征经过LASSO回归与十折交叉验证选择最优lambda值0.023，最终筛选出6个最优形态学特征，根据这六大形态学特征构建SVM、KNN、LR机器学习模型，其曲线下面积（AUC）分别为 0.73（95% CI 0.47~0.98）、0.80（95% CI 0.72~0.87）、0.75（95% CI 0.61~0.89）。基于高血压这一临床特征和形态学特征构建的SVM1、KNN1和LR1模型的AUC分别为0.79（95% CI 0.67~0.93）、0.85（95% CI 0.79~0.92）、0.83（95% CI 0.72~0.95）。结论 本研究基于影像组学技术自动提取颅内动脉瘤的形态学特征，并构建了六种机器学习模型，能准确识别动脉瘤的状态，从而可以尽早对具有高破裂风险的颅内动脉瘤进行干预，具有重要的临床指导意义。
Abstract:
Objective To explore the effectiveness of machine learning models based on clinico-radiomics features in predicting the risk of intracranial aneurysm rupture. Methods The clinical data of 153 patients with intracranial aneurysms admitted to our hospital from January 2019 to December 2022 were retrospectively analyzed. Univariate analysis was used to screen the clinical features affecting aneurysm rupture. DICOM format image data were collected, and the 3D Slicer software was used to reconstruct and segment the parent artery. The radiomics plug-in was used to extract the morphological features of the aneurysm, and the LASSO algorithm was used to screen the important morphological features affecting aneurysm rupture. Machine learning models based on clinical and morphological features were constructed, and the area under the curve (AUC), accuracy, precision, sensitivity, and specificity of each model in predicting the risk of intracranial aneurysm rupture were calculated. Results Of these 153 patients, 43 patients had ruptured intracranial aneurysms (ruptured group) and 110 had unruptured intracranial aneurysms (unruptured group). Compared with the unruptured group, the proportion of patients with hypertension was significantly higher (P<0.05), the elongation and sphericity of the aneurysm morphological features were significantly reduced (P<0.05), while the minimum axial diameter, maximum axial diameter, maximum coronal diameter, maximum three-dimensional diameter, grid volume, surface area, surface area volume ratio, and voxel volume of the aneurysm were significantly increased in the ruptured group (P<0.05). The AUC of SVM, KNN and LR models based on the six morphological features were 0.73 (95% CI 0.47~0.98), 0.80 (95% CI 0.72~0.87), 0.75 (95% CI 0.61~0.89), respectively. The AUC of SVM1, KNN1 and LR1 models based on the clinical feature of hypertension and morphological features were 0.79 (95% CI 0.67~0.93), 0.85 (95% CI 0.79~0.92), 0.83 (95% CI 0.72~0.95), respectively. Conclusions The morphological features of intracranial aneurysms were automatically extracted based on radiomics technology, and six machine learning models were constructed, which can accurately identify the status of aneurysms, so as to intervene as early as possible for intracranial aneurysms with high risk of rupture, which has important clinical significance.

参考文献/References:

[1] VLAK MH, ALGRA A, BRANDENBURG R, et al. Prevalence of unruptured intracranial aneurysms, with emphasis on sex, age, comorbidity, country, and time period: a systematic review and meta-analysis [J]. Lancet Neurol, 2011, 10(7): 626-636.
[2] 耿介文，翟晓东，吉　喆，等. 中国颅内未破裂动脉瘤诊疗指南2021[J]. 中国脑血管病杂志，2021，18（9）：634-664.
[3] GREVING JP, WERMER MJ, BROWN RD, et al. Development of the PHASES score for prediction of risk of rupture of intracranial aneurysms: a pooled analysis of six prospective cohort studies [J]. Lancet Neurol, 2014, 13(1): 59-66.
[4] MORITA A, KIRINO T, HASHI K, et al. The natural course of unruptured cerebral aneurysms in a Japanese cohort [J]. N Engl J Med, 2012, 366(26): 2474-2482.
[5] MA D, TREMMEL M, PALUCH RA, et al. Size ratio for clinical assessment of intracranial aneurysm rupture risk [J]. Neurol Res, 2010, 32(5): 482-486.
[6] ZHANG J, CAN A, MUKUNDAN S, et al. Morphological variables associated with ruptured middle cerebral artery aneurysms [J]. Neurosurgery, 2019, 85(1): 75-83.
[7] GILLIES RJ, KINAHAN PE, HRICAK H. Radiomics: images are more than pictures, they are data [J]. Radiology, 2016, 278(2): 563-577.
[8] CHOI RY, COYNER AS, KALPATHY-CRAMER J, et al. Introduction to machine learning, neural networks, and deep learning [J]. Transl Vis Sci Technol, 2020, 9(2): 14.
[9] ZHU W, LI W, TIAN Z, et al. Stability assessment of intracranial aneurysms using machine learning based on clinical and morphological features [J]. Transl Stroke Res, 2020, 11(6): 1287-1295.
[10] BIJLENGA P, GONDAR R, SCHILLING S, et al. PHASES score for the management of intracranial aneurysm: a cross-sectional population-based retrospective study [J]. Stroke, 2017, 48(8): 2105-2112.
[11] TANIOKA S, ISHIDA F, YAMAMOTO A, et al. Machine learning classification of cerebral aneurysm rupture status with morphologic variables and hemodynamic parameters [J]. Radiol Artif Intell, 2020, 2(1): e190077.
[12] ZHU W, LI W, TIAN Z, et al. Stability assessment of intra-cranial aneurysms using machine learning based on clinical and morphological features [J]. Transl Stroke Res. 2020, 11(6): 1287-1295.
[13] SILVA MA, PATEL J, KAVOURIDIS V, et al. Machine learning models can detect aneurysm rupture and identify clinical features associated with rupture [J]. World Neuro-surg, 2019, 131: 46-51.

相似文献/References:

[1]曾　春　张施远　蒋永明.复合手术与常规夹闭术治疗颈内动脉后交通动脉破裂动脉瘤的疗效对比研究[J].中国临床神经外科杂志,2016,(05):257.[doi:10.13798/j.issn.1009-153X.2016.05.001]
　ZENG Chun,ZHANG Shi-yuan,JIANG Yong-ming.Analysis of the effects of hybrid surgery on posterior communicating artery aneurysms: a comparison with conventional craniotomy[J].,2016,(09):257.[doi:10.13798/j.issn.1009-153X.2016.05.001]
[2]周有东　敖祥生　刘汉东　王志勇　朱耀祖　周达全.支架辅助弹簧圈和单纯弹簧圈栓塞治疗颅内动脉瘤疗效及安全性的Meta分析[J].中国临床神经外科杂志,2016,(05):274.[doi:10.13798/j.issn.1009-153X.2016.05.006]
　ZHOU You-dong,AO Xiang-sheng,LIU Han-dong,et al.A mata-analysis of curative effects of stent-assisted coiling and simple coiling on intracranial aneurysms and their safety[J].,2016,(09):274.[doi:10.13798/j.issn.1009-153X.2016.05.006]
[3]赵　平　王雄伟　汪　雷　马金阳.显微手术和血管内栓塞术治疗颅内动脉瘤的疗效分析[J].中国临床神经外科杂志,2016,(05):281.[doi:10.13798/j.issn.1009-153X.2016.05.008]
　ZHAO Ping,WANG Xiong-wei,WANG Lei,et al.Effect of endovascular treatment on intracranial aneurysms: a comparison with microsurgery[J].,2016,(09):281.[doi:10.13798/j.issn.1009-153X.2016.05.008]
[4]李朝晖　王玉贵　张孟超 综述　赵兴利 审校.CT灌注成像在动脉瘤性蛛网膜下腔出血中的 临床应用进展[J].中国临床神经外科杂志,2016,(05):313.[doi:10.13798/j.issn.1009-153X.2016.05.021]
[5]刘福德　邓剑平　张　涛　陈　虎　李　帅　王俭博　曲友直　赵振伟.椎-基底动脉夹层动脉瘤的血管内治疗（附48例分析）[J].中国临床神经外科杂志,2016,(06):353.[doi:10.13798/j.issn.1009-153X.2016.06.011]
　LIU Fu-de,DENG Jian-ping,ZHANG Tao,et al.Endovascular treatment of vertebrobasilar artery dissecting aneurysms (report of 48 cases)[J].,2016,(09):353.[doi:10.13798/j.issn.1009-153X.2016.06.011]
[6]吴有志　罗良生　张　健　陈　骅　吴　鸣　刘　振　史　岩.前循环破裂动脉瘤伴脑内血肿的 早期显微手术治疗[J].中国临床神经外科杂志,2016,(06):376.[doi:10.13798/j.issn.1009-153X.2016.06.020]
[7]潘　力　刘　鹏　秦　杰　杨　柳　安学峰　吴　虓　杨　铭　徐国政　马廉亭.Y型支架辅助技术在颅内动脉分叉部宽颈动脉瘤治疗中的应用[J].中国临床神经外科杂志,2016,(07):394.[doi:10.13798/j.issn.1009-153X.2016.07.002]
　PAN Li,LIU Peng,QIN Jie,et al.Application of Y -configuration stent technique to treatment of intracranial wide-neck bifurcation aneurysms[J].,2016,(09):394.[doi:10.13798/j.issn.1009-153X.2016.07.002]
[8]周　量　刘　丛　张昌伟.早期栓塞术治疗动脉瘤性蛛网膜下腔出血的疗效[J].中国临床神经外科杂志,2016,(07):410.[doi:10.13798/j.issn.1009-153X.2016.07.007]
　ZHOU Liang,LIU Cong,ZHANG Chang-wei..Curative effect of early endovascular embolization on patients with ruptured intracranial aneurysms[J].,2016,(09):410.[doi:10.13798/j.issn.1009-153X.2016.07.007]
[9]张　莹　荆林凯　张倩倩　刘　健　张义森　王　坤　穆士卿　杨新健.颈内动脉后交通动脉小动脉瘤破裂危险因素分析[J].中国临床神经外科杂志,2016,(03):138.[doi:10.13798/j.issn.1009-153X.2016.03.002]
　ZHANG Ying,JING Lin-kai,ZHANG Qian-qian,et al.Analysis of risk factors related to rupture of small (diameter ≤5 mm) posterior communicating artery aneurysms[J].,2016,(09):138.[doi:10.13798/j.issn.1009-153X.2016.03.002]
[10]朱作磊 综述　罗良生 审校.动脉瘤性蛛网膜下腔出血的预后相关因素研究进展[J].中国临床神经外科杂志,2016,(03):183.[doi:10.13798/j.issn.1009-153X.2016.03.020]

备注/Memo

备注/Memo:
（2023-04-10收稿，2023-08-08修回）
基金项目：苏州市科技发展计划（医疗卫生科技创新）项目（SKY2021054）
通讯作者：丁志良，E-mail：zlding1970@163.com
更新日期/Last Update: 2022-09-30