Page 128 - 南京医科大学学报自然科学版
P. 128

第42卷第8期
               ·1170 ·                           南 京    医 科 大 学 学         报                        2022年8月


              其他特征筛选法所建模型的差异,后期需进一步选                                 Soc,2017,2017:3012-3015
              择不同的方法比较模型的效能,以选择最优效能,                            [10] ALBERS G W,MARKS M P,KEMP S,et al. Thrombecto⁃
              适合临床应用的模型。                                             my for stroke at 6 to 16 hours with selection by perfusion
                  综上所述,基于机器学习技术对DWI和PWI图                             imaging[J]. N Engl J Med,2018,378(8):708-718
                                                                [11] RAOULT H,LASSALLE M V,PARAT B,et al. Dwi ⁃
              像进行预测学习,能够较为准确的预测 AIS 机械取
                                                                     based algorithm to predict disability in patients treated
              栓治疗后预后,为临床后续干预治疗及康复提供依
                                                                     with thrombectomy for acute stroke[J]. AJNR Am J Neu⁃
              据。
                                                                     roradiol,2020,41(2):274-279
             [参考文献]                                             [12] MUNDIYANAPURATH S,RINGLEB P A,DIATSCHUK
                                                                     S,et al. Time⁃dependent parameter of perfusion imaging
             [1] VAN DEN BERG L A,DIJKGRAAF M G,BERKHEMER
                                                                     as independent predictor of clinical outcome in symptom⁃
                   O A,et al. Two⁃year outcome after endovascular treatment
                                                                     atic carotid artery stenosis[J]. BMC Neurol,2016,16:50
                   for acute ischemic stroke[J]. N Engl J Med,2017,376
                                                                [13] PARSONS M W,CHRISTENSEN S,MCELDUFF P,et al.
                  (14):1341-1349
             [2] 龚鹏宇,周俊山,龚亚驰,等. 轻度急性缺血性卒中早期                          Echoplanar Imaging Thrombolytic Evaluation Trial(EPI⁃
                   神经功能恶化的风险因素及列线图预测模型的构建                            THET)Investigators. Pretreatment diffusion ⁃ and perfu⁃
                  [J]. 南京医科大学学报(自然科学版),2021,41(7):                   sion⁃MR lesion volumes have a crucial influence on clini⁃
                                                                     cal response to stroke thrombolysis[J]. J Cereb Blood
                   1039-1043
             [3] JIANG L,PENG M,CHEN H,et al. Diffusion⁃weighted     Flow Metab,2010,30(6):1214-1225
                   imaging(DWI)ischemic volume is related to FLAIR hy⁃  [14] GIGER M L. Machine learning in medical imaging[J]. J
                   perintensity⁃DWI mismatch and functional outcome after  Am Coll Radiol,2018,15(3 Pt B):512-520
                   endovascular therapy[J]. Quant Imaging Med Surg,2020,  [15] 刘  娜,隋庆兰,刘学军,等. 增强 MRI 影像组学在高
                                                                     级别胶质瘤 IDH 1 基因型预测方面的价值[J].中华放
                   10(2):356-367
             [4] WOLMAN D N,IV M,WINTERMARK M,et al. Can dif⁃        射学杂志,2020,54(5):445-449
                                                                [16] 徐青青,单文莉,朱        艳,等. 基于 CT 影像组学对孤立
                   fusion⁃ and perfusion⁃weighted imaging alone accurately
                   triage anterior circulation acute ischemic stroke patients  性肺结节性质分类的预测效能[J]. 南京医科大学学报
                   to endovascular therapy?[J]. J Neurointerv Surg,2018,10  (自然科学版),2021,41(4):617-623
                  (12):1132-1136                                [17] HEO J,YOON J G,PARK H,et al. Machine learning ⁃
             [5] SIRSAT M S,FERMÉ E,CÂMARA J. Machine learning       based model for prediction of outcomes in acute stroke
                   for brain stroke:a review[J]. J Stroke Cerebrovasc Dis,  [J]. Stroke,2019,50(5):1263-1265
                   2020,29(10):105162                           [18] 上官艺,王     孟,王春娟,等. 基于机器学习的缺血性卒
             [6] LI X,PAN X,JIANG C,et al. Predicting 6⁃month unfavor⁃  中功能预后预测模型研究[J]. 中国卒中杂志,2021,16
                   able outcome of acute ischemic stroke using machine  (9):895-900
                   learning[J]. Front Neurol,2020,11:539509     [19] KIM Y C,LEE J E,YU I,et al. Evaluation of diffusion le⁃
             [7] CHENG C,HUA Z C. Lasso peptides:heterologous pro⁃   sion volume measurements in acute ischemic stroke using
                   duction and potential medical application[J]. Front Bio⁃  encoder⁃decoder convolutional network[J]. Stroke,2019,
                   eng Biotechnol,2020,8:571165                      50(6):1444-1451
             [8] MUNDIYANAPURATH S,DIATSCHUK S,LOEBEL S,        [20] XIE Y,JIANG B,GONG E,et al. JOURNAL CLUB:Use
                   et al. Outcome of patients with proximal vessel occlusion  of gradient boosting machine learning to predict patient
                   of the anterior circulation and DWI⁃PWI mismatch is time  outcome in acute ischemic stroke on the basis of imaging,
                   ⁃dependent[J]. Eur J Radiol,2017,91:82-87         demographic,and clinical information[J]. AJR Am J
             [9] VUPPUTURI A,ASHWAL S,TSAO B,et al. MRI based        Roentgenol,2019,212(1):44-51
                   objective ischemic core⁃penumbra quantification in adult               [收稿日期] 2022-03-30
                   clinical stroke[J]. Annu Int Conf IEEE Eng Med Biol                         (本文编辑:唐      震)
   123   124   125   126   127   128   129   130   131   132   133