基于深度学习实现维吾尔族青少年左手腕关节骨龄自动化评估  

Automated Assessment for Bone Age of Left Wrist Joint in Uyghur Teenagers by Deep Learning

作　　者：胡婷鸿 火忠[3] 刘太昂 王飞[3] 万雷 汪茂文 陈腾 王亚辉 HU Ting-hong;HUO Zhong;LIU Tai-ang;WANG Fei;WAN Lei;WANG Mao-wen;CHEN Teng;WANG Ya-hui(Shanghai Key Laboratory of Forensic Medicine, Shanghai Forensic Service Platform, Academy of Forensic Science, Shanghai 200063,China;Department of Forensic Science, Xi’an Jiaotong University Health Science Center, Xi’an 710061, China;People’s Hospital of Xinjiang Uygur Autonomous Region, Urumqi 830000, China;Shanghai Fanyang Information Technology Co., LTD., Shanghai 200444, China)

机构地区：[1]司法鉴定科学研究院,上海市法医学重点实验室,上海市司法鉴定专业技术服务平台,上海200063 [2]西安交通大学医学部法医学院,陕西西安710061 [3]新疆维吾尔自治区人民医院,新疆乌鲁木齐830000 [4]上海帆阳信息科技有限公司,上海200444

出　　处：《法医学杂志》2018年第1期27-32,共6页Journal of Forensic Medicine

基　　金：国家自然科学基金资助项目(81571859,81102305,81401559);上海市法医学重点实验室资助项目(17DZ2273200);上海市司法鉴定专业技术服务平台资助项目(16DZ2290900);上海市法医学重点实验室开放基金资助项目(KF1706)

摘　　要：目的将深度学习运用于维吾尔族青少年左手腕关节数字化X线摄影(digital radiography,DR)图像识别中,实现骨龄评估的自动化,探索该方法在法医骨龄鉴定中的应用价值。方法在我国新疆维吾尔自治区采集13.0~19.0岁维吾尔族男性青少年245例、女性青少年227例左手腕关节DR图像,将预处理后的图像作为研究对象,将AlexNet作为图像识别的回归模型。在上述总样本中分别选取男、女性60%左手腕关节DR图像样本作为网络训练集,10%的样本作为验证集,余30%作为测试集,获取与样本真实年龄误差范围分别在±1.0岁、±0.7岁以内的图像识别准确率。结果深度学习的内测结果:误差范围在±1.0岁及±0.7岁以内的网络训练集准确率,男性分别为81.4%和75.6%,女性分别为80.5%和74.8%。误差范围在±1.0岁及±0.7岁以内的测试集准确率,男性分别为79.5%和71.2%,女性分别为79.4%和66.2%。结论青少年左手腕关节骨龄研究与深度学习相结合,具有较高的准确性及较好的可行性,为躯体其余骨关节的骨龄自动化评估体系奠定研究基础。Objective To realize the automated bone age assessment by applying deep learning to digital radiography(DR)image recognition of left wrist joint in Uyghur teenagers,and explore its practical application value in forensic medicine bone age assessment.Methods The X-ray films of left wrist joint after pretreatment,which were taken from245male and227female Uyghur nationality teenagers in Uygur Autonomous Region aged from13.0to19.0years old,were chosen as subjects.And AlexNet was as a regression model of image recognition.From the total samples above,60%of male and female DR images of left wrist joint were selected as net train set,and10%of samples were selected as validation set.As test set,the rest30%were used to obtain the image recognition accuracy with an error range in±1.0and±0.7age respectively,compared to the real age.Results The modelling results of deep learning algorithm showed that when the error range was in±1.0and±0.7age respectively,the accuracy of the net train set was81.4%and75.6%in male,and80.5%and74.8%in female,respectively.When the error range was in±1.0and±0.7age respectively,the accuracy of the test set was79.5%and71.2%in male,and79.4%and66.2%in female,respectively.Conclusion The combination of bone age research on teenagers’left wrist joint and deep learning,which has high accuracy and good feasibility,can be the research basis of bone age automatic assessment system for the rest joints of body.

关 键 词：法医人类学 年龄测定 骨骼 腕关节 体层摄影术 X线 图像识别 深度学习 维吾尔族 青少年 

分 类 号：DF795.6[医药卫生—法医学]