深度学习模型在儿童TW3法骨龄评估的临床效能分析  

The clinical efficacy of deep learning model in bone age ssessment of children using TW3 method

作　　者：程梦 钱琦[1] 高铖铖 田曼曼[1] 励杨晟 林敏 Cheng Meng

机构地区：[1]浙江中医药大学附属第三医院,310005 [2]浙江大学医学院附属杭州市第一人民医院,310006 [3]浙江中医药大学人文与管理学院,311402

出　　处：《浙江临床医学》2023年第1期25-28,共4页Zhejiang Clinical Medical Journal

基　　金：浙江省医药卫生科技计划项目(2020PY014)。

摘　　要：目的探讨深度学习模型在儿童应用TW3法进行骨龄评估的临床效能。方法回顾性收集180例儿童左手X线片。评估基于Tanner-Whitehouse III(TW3)法的掌指骨和腕骨骨龄。以3位高年资主任医师的评估均值为金标准组,计算并比较深度学习模型(模型组)及2位低年资医师(医师组,分别记作医师1、医师2)与金标准组评估时间差异,掌指骨骨龄、腕骨骨龄的均方误差(MSE)及平均绝对误差(MAE);采用组内相关系数(ICC)分析模型组和医师组与金标准组结果一致性。结果骨龄评估所用时间模型组明显少于医师组(P<0.05)。掌指骨骨龄评估,模型组和医师1、医师2与金标准组相比MSE、MAE差异有统计学意义(P<0.05)。腕骨骨龄评估,模型组和医师1、医师2与金标准组相比MSE、MAE差异有统计学意义(P<0.05)。掌指骨骨龄评估,模型组与金标准组ICC为0.988,医师1、医师2与金标准组的ICC分别为0.986、0.977。腕骨骨龄评估,模型组与金标准组ICC为0.971,医师1、医师2与金标准组ICC分别为0.970、0.953。结论对于儿童,应用TW3法深度学习模型在评估掌指骨和腕骨骨龄有临床价值。Objective To explore the clinical efficacy of deep learning model in bone age assessment(BAA)of children using TW3 method,based on artificial intelligence(AI)system.Methods Totally 180 left hand and wrist radiographs of children aged 5~13 years in Zhejiang Province were retrospectively studied.Tanner-Whitehouse III(TW3)method based on radius,ulna and short bone(RUS)and carpal bone ages were assessed.The average value of bone age estimated by three chief radiologists was taken as reference standard group,two junior radiologists(denoted as doctor 1,2)as doctor group,and AI system as model group.The BAA time,mean square error(MSE)and mean absolute error(MAE)of BAA relative to reference standard group were calculated and compared among three groups.The intraclass correlation coefficient(ICC)was used to compare the consistency of the results among the model group and the doctor group relative to reference standard group.Results BAA time of the model group was significantly shorter than that of the doctor group(P<0.05).There were significant differences of TW3 RUS's BAA MSE and MAE between the model group and doctor 1,2(all P<0.05).There were significant differences of carpal's BAA MSE and MAE between model group and doctor 1,2(all P<0.05).ICC between RUS's BAA of the model group and the reference standard group was 0.988,ICC between the doctor group(doctor 1,2)and the reference standard group was 0.986 and 0.977,respectively.ICC between carpal's BAA of the model group and the reference standard group was 0.971,ICC between the doctor group(doctor 1,2)and the reference standard group was 0.970 and 0.953,respectively.Conclusion The deep learning model using TW3 method has clinical value in evaluating the RUS and carpal bone ages of children in Zhejiang Province.

关 键 词：骨龄 深度学习 TW3法 儿童 

分 类 号：R68[医药卫生—骨科学]