医用X线图像输入分辨率对转子间骨折识别YOLO网络效果的影响  

Impact of input image resolution in medical X-ray images on effectiveness of YOLO network for recognition of intertrochanteric fractures

作　　者：刘学思 杜振伟 聂瑞 李翔 韩海平 张乐天 李阳 张和华 LIU Xuesi;DU Zhenwei;NIE Rui;LI Xiang;HAN Haiping;ZHANG Letian;LI Yang;ZHANG Hehua(Department of Medical Engineering,State Key Laboratory of Trauma and Chemical Poisoning,Army Medical Center of PLA,Chongqing,400042;Department of Information,First Affiliated Hospital,Army Medical University(Third Military Medical University),Chongqing,400038;Health Company of Troop 65547 of PLA,Anshan,Liaoning Province,114000,China;Department of Radiology,State Key Laboratory of Trauma and Chemical Poisoning,Army Medical Center of PLA,Chongqing,400042;Division of Trauma and War Injury,State Key Laboratory of Trauma and Chemical Poisoning,Army Medical Center of PLA,Chongqing,400042)

机构地区：[1]陆军特色医学中心医学:医学工程科,重庆400042 [2]陆军军医大学(第三军医大学)第一附属医院信息科,重庆400038 [3]中国人民解放军第65547部队卫生连,辽宁鞍山114000 [4]陆军特色医学中心医学放射科,重庆400042 [5]陆军特色医学中心医学战创伤医学中心,创伤与化学中毒全国重点室验室,重庆400042

出　　处：《陆军军医大学学报》2023年第22期2327-2333,共7页Journal of Army Medical University

基　　金：重庆市科技创新与应用发展专项重点项目(CSTB2021TIAD-KPX0074);陆军军医大学优秀人才库重点扶持对象项目(B-3267)。

摘　　要：目的以股骨转子间骨折数据集为研究对象,探究医用X线图像输入分辨率的改变对YOLO系列目标检测网络识别效果的影响。方法收集陆军特色医学中心2017-2022年间患转子间骨折患者的X线正位数据,按排除标准排除后共保留426例患者,847幅图像。结合2018版Arbeitsgemeinschaft für Osteosynthesefragen/Orthopaedic Trauma Association(AO/OTA)分型指南和临床实际发病率,将转子间骨折重新整理划分为A1.2、A1.3、A2.2、A2.3、A3等5个亚组,并按8∶1∶1的比例划分为训练集(678幅)、验证集(84幅)和测试集(85幅),每次试验中保持严格一致。将YOLOX-Swin-Transformer、YOLOX、YOLOv5、YOLOv4目标识别网络图像输入大小分别设置为8种常用的分辨率,使用从头训练和迁移学习方法分别训练转子间骨折数据集,记录训练耗时,利用测试集进行模型测试并记录评价指标。采用SPSS20.0统计软件回归分析对训练耗时和均值平均精度(mean average precision,mAP)做曲线拟合,利用频数统计功能对各输入分辨率下评价指标评级为优的频数进行统计,以此判断最优区间。结果图像输入分辨率与各网络训练耗时呈正相关,所有P<0.05,线性回归分析有统计学意义。图像输入分辨率与网络mAP均值二次曲线拟合R^(2)=0.834,P=0.011,说明曲线拟合优度较好,回归分析有统计学意义。当图像输入分辨率在480×480、576×576、640×640范围内时,数据评级为优出现的频率最高(占42.86%)。结论训练耗时随分辨率的增加而增加。在不改变网络结构的前提下采用YOLO系列网络做医学图像识别的下游任务时,要取得最佳的识别效果,图像输入分辨率应控制在480×480、576×576、640×640范围之内。Objective To explore the effect of various input resolution of X-ray images on the performance of the You Only Look Once(YOLO)network in recognition of intertrochanteric fractures.MethodsX-ray anteroposterior data of the patients with intertrochanteric fractures admitted in Army Medical Center of PLA from 2017 to 2022 were collected,and finally,426 patients and 847 images were retained after exclusion criteria.Based on the 2018 guideline of Arbeitsgemeinschaft für Osteosynthesefragen/Orthopaedic Trauma Association(AO/OTA)and actual clinical incidence,these intertrochanteric fractures were reclassified into grades A1.2/A1.3/A2.2/A2.3/A3,and the X-ray images were assigned into training set(678 images),validation set(84 images),and test set(85 images)in a ratio of 8∶1∶1 in order to maintain strict consistency across each experiment.Eight common resolutions were set as input size for YOLOX-Swin-Transformer,YOLOX,YOLOv5,and YOLOv4 object detection networks.The training set was trained using both training from scratch and transfer learning.The training time was recorded,the test set was used to test the model,and evaluation metrics was recorded.SPSS20.0 statistical software was employed for statistical analysis.Regression analysis was applied to test curve fitting of training time and mean average precision(mAP)values.Frequency statistics function was performed to count the frequencies of evaluation indicators rated as excellent at each input resolution in order to determine the optimal range.Results The image input resolution was positively correlated with the training time of various networks,with all P-values<0.05,showing statistical significance by linear regression analysis.The quadratic curve fitting of the image input resolution and the mAP mean value of the network resulted in an R^(2)=0.834(R^(2)>0.5)and P=0.011(P<0.05),indicating a good fit of the curve and statistical significance in the regression analysis.When the input image resolution was in a range of 480×480,576×576,640×640,the frequency of opt

关 键 词：转子间骨折 AO/OTA分型 目标检测 X线 YOLO 

分 类 号：R319[医药卫生—基础医学] R683.42R812