摘要
目的通过深度残差神经网络(ResNet38)对腹部立位平片进行学习,从中检出小肠梗阻。方法本研究使用的训练集和测试集数据来源于西安交通大学第一附属医院及陕西省核工业二一五医院,独立验证集数据来源于陕西省核工业二一五医院。由2名经验丰富的影像科诊断医师依据腹部立位平片小肠梗阻的4种征象分别对来源于2家医院的3298张腹部立位平片进行分类,意见不一致时讨论达到共识,其中小肠梗阻569例(17.3%),非小肠梗阻2729例(82.7%)。对2组数据采用完全随机分组的方法组成训练集2305张和测试集993张(训练集∶测试集=2.3∶1),其中训练集小肠梗阻405例(17.6%),非小肠梗阻1900例(82.4%);测试集小肠梗阻164例(16.5%),非小肠梗阻829例(83.5%)。训练集和测试集小肠梗阻的诊断均以有丰富经验的影像科医师的评判为标准。验证集共861张腹部立位平片,其中小肠梗阻99例(11.5%),非小肠梗阻762例(88.5%),以手术结果及临床诊断为金标准。本研究使用ImageNet 2012年大规模视觉识别挑战赛数据集(ILSVRC2012)对深度残差神经网络(ResNet38)进行预训练;用训练集数据对深度残差神经网络(ResNet38)再训练建立诊断模型;测试集主要用于学习算法过程中,调整算法的参数来修正网络,从而使得网络模型效能更优。结果本研究开发的小肠梗阻诊断模型在测试集上敏感性为84.1%,特异性为65.0%,受试者工作特征曲线(receiver operating characteristic,ROC)曲线下面积(area under the curve,AUC)为0.83(95%CI为0.82~0.92);在验证集上的敏感性为89.9%,特异性为68.0%,ROC曲线下面积为0.87(95%CI为0.82~0.92),AUC为0.87。结论对深度残差神经网络进行有限的数据训练,可以获得一个有效的用于在腹部立位平片上检出小肠梗阻的诊断模型。
Objective To study whether a deep residual neural network can detect small bowel obstruction patterns on upright abdominal radiographs.Methods The data of training set and test set used in this study were obtained from The First Affiliated Hospital of Xi'an Jiaotong University and No.215 Hospital of Shaanxi Nuclear Industry;the data of validation set came from No.215 Hospital of Shaanxi Nuclear Industry.Totally 3298 clinical upright abdominal radiographs obtained from two hospitals were classified into obstructive and non-obstructive categories independently by two radiologists on the basis of the four signs on upright abdominal radiographs,who discussed and reached consensus when disagreements arose.Among them,569(17.3%)images were found to be consistent with small bowel obstruction,and 2729(82.7%)images had no small bowel obstruction.A total of 2305 training sets and 993 test sets(training set:test set=2.3∶1)were composed of data from the two groups,including 405 cases(17.6%)of small bowel obstruction,1900 cases(82.4%)of non-small bowel obstruction,164 cases(16.5%)of small bowel obstruction,and 829 cases(83.5%)of non-small bowel obstruction.The diagnosis of small bowel obstruction in training and testing sets was based on experienced radiologists evaluation.Totally 861 abdominal upright abdominal radiographs constituted the validation set(99 with small bowel obstruction and 762 with no small bowel obstruction);the surgical results and clinical diagnosis were set as the gold standard.In this study,the image 2012 large-scale visual recognition challenge data set(ILSVRC2012)was used for pre-training the deep residual neural network(ResNet38).The retraining of deep residual network(ResNet38)with training set data was used to establish the diagnostic model.The test set was mainly used in the learning algorithm process to adjust the algorithm parameters to modify the network,so as to make the network model more efficient.Results After training,the deep residual neural network achieved an AUC of 0.83 on the test set(95%CI 0.82-0.92).The sensitivity of the system for small bowel obstruction was 84.1%,with a specificity of 65.0%.And on validation set it achieved an AUC of 0.87(95%CI 0.82-0.92),the sensitivity of the system for small bowel obstruction was 89.9%,with a specificity of 68.0%.Conclusion Transfer learning with deep residual neural network may be used to train a detector for small bowel obstruction on upright abdominal radiographs even with limited training data.
作者
沈远望
李贤军
刘哲
尚进
孙亲利
刘尼军
曹盼
宋春晓
杨健
SHEN Yuan-wang;LI Xian-jun;LIU Zhe;SHANG Jin;SUN Qin-li;LIU Ni-jun;CAO Pan;SONG Chun-xiao;YANG Jian(Department of Medical Imaging,The First Affiliated Hospital of Xi'an Jiaotong University,Xi'an 710061;Department of Medical Imaging,No.215 Hospital of Shaanxi Nuclear Industry,Xianyang 712000;Department of Medical Imaging,Shaanxi Provincial Tuberculosis Hospital,Xi'an 710100;AccuRadNetwork&Technology Co.,Ltd.,Xi'an 710075,China)
出处
《西安交通大学学报（医学版）》
CAS
CSCD
北大核心
2020年第1期102-107,共6页
Journal of Xi’an Jiaotong University（Medical Sciences）
基金
国家重点研发计划项目(No.2016YFC0100300)~~
关键词
腹部平片
数字化X线摄影
小肠梗阻
残差神经网络
深度学习
abdominal plain film
digital radiography
small bowel obstruction
residual neural network
deep learning
