[1]陈宇航　林中松　李新瑜　张　亮　李小潘　郝晓伟　程小兵　张鸿日.基于迁移学习算法开发的人工智能脑积水诊断模型的临床应用[J].中国临床神经外科杂志,2021,26(05):349-351.[doi:10.13798/j.issn.1009-153X.2021.05.010]
　CHEN Yu-hang,LIN Zhong-song,LI Xin-yu,et al.Application of artificial intelligence model established based on transfer learning algorithm in diagnosis of hydrocephalus[J].,2021,26(05):349-351.[doi:10.13798/j.issn.1009-153X.2021.05.010]
点击复制

基于迁移学习算法开发的人工智能脑积水诊断模型的临床应用()
分享到：

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

卷:
26
期数:
2021年05期
页码:
349-351
栏目:
论著
出版日期:
2021-05-25

文章信息/Info

Title:
Application of artificial intelligence model established based on transfer learning algorithm in diagnosis of hydrocephalus
文章编号:
1009-153X（2021）05-0349-03
作者:
陈宇航　林中松　李新瑜　张　亮　李小潘　郝晓伟　程小兵　张鸿日
471000 河南洛阳，河南科技大学第一附属医院神经外科（陈宇航、郝晓伟、程小兵、张鸿日），影像中心（李新瑜、李小潘）；201203 上海，上海纳凝微信息科技有限公司（林中松、张　亮）
Author(s):
CHEN Yu-hang1;  LIN Zhong-song2;  LI Xin-yu3;  ZHANG Liang2;  LI Xiao-pan3;  HAO Xiao-wei1;  CHENG Xiao-bing1;  ZHANG Hong-ri1. 1.
Department of Neurosurgery, First Affiliated Hospital, Henan University of Science and Technology, Luoyang 471000, China; 2.Shanghai Nanoperception Technology Co. LTD, Shanghai 201203, China; 3. Department of Radiology, First Affiliated Hospital, Henan Un
关键词:
脑积水; CT影像; 诊断; 人工智能; 迁移学习算法
Keywords:
Hydrocephalus;  Artificial intelligence;  Transfer learning;  Diagnosis
分类号:
R 742.7
DOI:
10.13798/j.issn.1009-153X.2021.05.010
文献标志码:
A
摘要:
目的 利用迁移学习算法设计、开发一种人工智能脑积水影像诊断工具，并评价其应用颅脑CT平扫影像诊断脑积水的效果。方法 收集河南科技大学第一附属医院正常成人及脑积水的颅脑CT影像DICOM数据各1 250例，按6∶2∶2随机分为训练集、验证集和测试集。应用Python开发标记工具对数据进行预处理并标注特征变量，建立自主研发的逐像素ForrestNet-CNN算法模型，并对特征变量进行提取和深度学习。选取影像学住院医师、主治医师、副主任医师各2名对250例脑积水和250例正常影像进行测试，并与人工智能模型的测试结果进行比较。结果 人工智能模型、住院医师、主治医师、副主任医师诊断脑积水的敏感性分别为94.4%、92.8%、95.2%和96.4%，特异性分别为93.6%、94.0%、96.0%和97.6%，准确率分别为94.0%、93.4%、95.6%和97.0%。虽然人工智能模型诊断脑积水的敏感性、特异性、准确率明显低于副主任医师（P<0.05），但是与住院医师、主治医师均无统计学差异（P<0.05）。结论 本研究开发的人工智能模型可有效地识别脑积水的CT影像特点，具有较高的准确率。
Abstract:
Objective To explore the value of an artificial intelligence (AI) model established based on the transfer learning algorithm in the diagnosis of hydrocephalus. Methods The DICOM data of brain CT images of 1 250 normal adults and 1 250 patients with hydrocephalus were collected, and then were randomly divided into training set, verification set and test set at the ratio of 6:2:2. The data were preprocessed and the characteristic variables were labeled using Python tool, and then the AI model, which was established by a self-developed pixel-by-pixel ForrestNet-CNN algorithm, was used to extract and study the characteristic variables. The diagnostic results of AI model were compared with the results of imaging resident, attending physician and deputy chief physician, respectively. Results The sensitivities of the AI model, resident physician, attending physician, and deputy chief physician in diagnosing hydrocephalus were 94.4%, 92.8%, 95.2%, and 96.4%; the specificities were 93.6%, 94.0%, 96.0%, and 97.6%; the accuracy rates were 94.0%, 93.4%, 95.6% and 97.0%, respectively. The sensitivity, specificity and accuracy rate in diagnosing hydrocephalus of the AI model were significantly lower than those of the deputy chief physician (P<0.05). There was no statistical differences in the sensitivity, specificity and accuracy rate in diagnosing hydrocephalus among the AI model, the resident physician and the attending physician (P>0.05). Conclusions The AI model developed in this study can effectively identify the CT image characteristics of hydrocephalus with high accuracy.

参考文献/References:

[1] Chatzidakis EM, Barlas G, Condilis N, et al. Brain CT scan indexes in the normal pressure hydrocephalus: predictive value in the outcome of patients and correlation to the clini-cal symptoms [J]. Ann Ital Chir, 2008,79: 353-362.
[2] Kermany DS, Goldbaum M, Cai W, et al. Identifying medicaldiagnoses and treatablediseases by image-based deep learning [J]. Cell, 2018, 172: 1122-1131.
[3] 张　鹏，徐欣楠，王洪伟，等. 基于深度学习的计算机辅助肺癌诊断方法[J]. 计算机辅助设计与图形学学报，2018，30：90-99.
[4] Azizpour H, Razavian AS, Sullivan J, et al. Factors of trans-ferability for a generic ConvNet representation [J]. IEEE T Pattern Anal, 2016, 38: 1790-1802.
[5] Evans WA. An encephalographic ratio for estimating ventri-cular enlargement and cerebral atrophy [J]. Arch Neurol Psychiatry, 1942, 47: 931-937.
[6] Kaya M, Hajimirza S. Using a novel transfer learning methodfor designing thin film solar cells with enhanced quantum efficiencies [J]. Sci Rep, 2019, 9: 5034.
[7] Tomycz LD, Hale AT, George TM. Emerging insights and new perspectives on the nature of hydrocephalus [J]. PediatrNeurosurg, 2017, 52: 361-368.
[8] Toma AK, Holl E, Kitchen ND, et al. Evans’index revisited:theneed for an alternative in normal pressure hydrocephalus[J]. Neurosurgery, 2011, 68: 939-944.
[9] Prevedello LM, Erdal BS, Ryu JL, et al. Automated critical test findings identification and online notification system using artificial intelligence in imaging [J]. Radiology, 2017, 285: 923-931.
[10] Lin WY, Chen CH, Tseng YJ, et al. Predicting poststrokeactivities of daily living through a machine learning based approach on initiating rehabilitation [J]. Int J MedInform, 2018, 111: 159-164.

相似文献/References:

[1]周　平　邓　燕　王　辉　罗　杰　胡钧涛　刘开军　李安荣　李新建.脊髓脊膜膨出合并脑积水分流手术时机的选择[J].中国临床神经外科杂志,2015,(11):692.[doi:10.13798/j.issn.1009-153X.2015.11.017]
[2]李光亮　赵向南　王方艺.创伤性蛛网膜下腔出血继发脑积水的临床分析[J].中国临床神经外科杂志,2016,(07):433.[doi:10.13798/j.issn.1009-153X.2016.07.015]
[3]娄元华　李小勇　陈红伟　潘栋超　解东成　刘东升.婴儿脑积水分流术后硬膜下积液的临床分析[J].中国临床神经外科杂志,2016,(04):245.[doi:10.13798/j.issn.1009-153X.2016.04.019]
[4]张家康　彭　飞　陈　鑫　侯　旭　张大明　刘震东　赵世光.后颅窝减压术治疗Chiari畸形Ⅰ型合并脑积水疗效的Meta分析[J].中国临床神经外科杂志,2016,(03):158.[doi:10.13798/j.issn.1009-153X.2016.03.009]
　ZHANG Jia-kang,PENG Fei,CHEN Xin,et al.Posterion fossa decompression for Chiari I malformation with hydrocephalus: a meta-analysis[J].,2016,(05):158.[doi:10.13798/j.issn.1009-153X.2016.03.009]
[5]云德波　张　逵　范润金　唐　辉　谯　飞　陈华轩.脑室-腹腔分流术后分流不畅的原因分析及处理[J].中国临床神经外科杂志,2016,(02):92.[doi:10.13798/j.issn.1009-153X.2016.02.011]
　YUN De-bo,ZHANG Kui,FAN Run-Jin,et al.Analysis of cause and countermeasures of poor diversion after ventriculoperitoneal shunt in patients with hydrocephalus[J].,2016,(05):92.[doi:10.13798/j.issn.1009-153X.2016.02.011]
[6]张建民.颅骨缺损合并脑积水及脑软化灶形成的手术治疗（附40例报告）[J].中国临床神经外科杂志,2016,(01):51.[doi:10.13798/j.issn.1009-153X.2016.01.019]
[7]申汉威　李俊卿　李红星　汪立刚　杨孔宾.颅内破裂动脉瘤夹闭术与血管内栓塞术后脑积水发生率的比较分析[J].中国临床神经外科杂志,2015,(03):140.[doi:10.13798/j.issn.1009-153X.2015.03.004]
　SHEN Han-wei,LI Jun-qing,LI Hong-xing,et al.Clinical study and Meta-analysis of incidences of hydrocephalus after microsurgery and endovascular embolization in patients with ruptured intracranial aneurysms[J].,2015,(05):140.[doi:10.13798/j.issn.1009-153X.2015.03.004]
[8]于　峰　王　路　门学忠　朱　冰　孙希炎.重度脑积水脑室-腹腔分流术后继发硬膜下血肿2例[J].中国临床神经外科杂志,2015,(03):192.[doi:10.13798/j.issn.1009-153X.2015.03.025]
[9]任瑞明　成文平　楚鹏飞　张　伟　李会兵　张　坤.脑室-腹腔分流术治疗脑积水48例[J].中国临床神经外科杂志,2015,(01):50.[doi:10.13798/j.issn.1009-153X.2015.01.017]
[10]孙　强.脑室端梗阻致脑室-腹腔分流术失败的原因及防治[J].中国临床神经外科杂志,2016,(10):620.[doi:10.13798/j.issn.1009-153X.2016.10.018]

备注/Memo

备注/Memo:
基金项目：河南省医学科技攻关计划省部共建项目（SB201901066）
通讯作者：张鸿日，E-mail：hongrizhang@126.com
更新日期/Last Update: 2021-05-25