人工智能软件在儿童发育相关疾病骨龄评估中的临床应用

中华养生保健 ›› 2023, Vol. 41 ›› Issue (22): 54-57.

人工智能软件在儿童发育相关疾病骨龄评估中的临床应用

邓成清¹, 孙安林², 唐宇轩², 张萍¹, 唐爱群^2,*

1.盐城市射阳县人民医院放射科,江苏盐城,224300;
2.盐城市射阳县人民医院儿科,江苏盐城,224300

出版日期:2023-11-16 发布日期:2023-11-08
通讯作者: ^*唐爱群,E-mail：190059059@qq.com。
作者简介:邓成清（1977—）,男,汉族,籍贯：江苏省盐城市,本科,主任医师,研究方向：医学影像。

Clinical Application of Artificial Intelligence Software in Bone Age Assessment of Developmental Related Diseases in Children

DENG Cheng-qing¹, SUN An-lin², TANG Yu-xuan², ZHANG Ping¹, TANG Ai-qun^2,*

1. Department of Radiology, Sheyang County People's Hospital, Yancheng City, Yancheng Jiangsu 224300, China;
2. Department of Pediatrics, Sheyang County, People's Hospital, Yancheng City, Yancheng Jiangsu 224300, China

Online:2023-11-16 Published:2023-11-08

摘要/Abstract

摘要： 目的探讨与分析人工智能软件在儿童发育相关疾病骨龄评估中的临床应用价值。方法选取2020年5月—2022年12月在江苏省盐城市射阳县人民医院诊治的84例发育相关疾病儿童作为研究对象,患儿都给予左手数字化X线检查,采用杭州深睿博联科技有限公司人工智能影像分析软件对数字化X线摄片进行骨龄评估,同时也由影像科住院医师与影像科主任医师进行阅片,以影像科主任医师的判断结果作为金标准,判定评估价值。结果人工智能阅片时间显著少于住院医师阅片与金标准阅片,差异有统计学意义（P<0.05）。住院医师阅片与金标准阅片时间比较,差异无统计学意义（P>0.05）。人工智能阅片误差与金标准阅片比较,差异无统计学意义（P>0.05）,住院医师阅片误差与金标准阅片比较,差异无统计学意义（P<0.05）。住院医师阅片、人工智能阅片、金标准阅片准确率分别为88.10%、97.62%、98.81%,人工智能阅片、金标准阅片的准确率显著高于住院医师阅片,差异有统计学意义（P<0.05）。人工智能阅片、金标准阅片的准确率比较,差异无统计学意义（P>0.05）。结论人工智能软件在儿童发育相关疾病骨龄评估中的临床应用能缩短阅片时间,提高对于骨龄判定的准确性,缩小阅片的误差。

关键词: 人工智能软件, 儿童发育相关疾病, 骨龄, 阅片时间, 准确性, 误差

Abstract: Objective To explore and analyse the clinical application values of artificial intelligence software in bone age assessment of developmental related diseases in children. Methods 84 cases of children with developmental related diseases diagnosed and treated at a certain hospital from May 2020 to December 2022 were selected as the research subjects. All children were underwent left hand digital X-ray examination, and bone age assessment were conducted using artificial intelligence imaging analysis software from Hangzhou Shenrui Bolian Technology Co., Ltd. The digital X-ray films were also reviewed by resident imaging physicians and chief imaging physicians. Used the chief physician of the imaging department as the gold standard to determine the evaluation value. Results AI reading time was significantly less than that of residents and gold standard （P<0.05）, and there was no significant difference between residents and gold standard （P>0.05）. There was no significant difference between the AI reading error （P>0.05）, and the resident and gold standard reading error （P<0.05）. The accuracy of resident reading, artificial intelligence reading and gold standard reading was 88.10%, 97.62% and 98.81% respectively. The accuracy of AI reading and gold standard reading was significantly higher than that of resident reading （P<0.05）, and there was no significant difference in the accuracy of AI reading and gold standard reading （P>0.05）. Conclusion The clinical application of artificial intelligence software in the assessment of bone age in children with developmental related diseases can shorten the reading time, improve the accuracy of bone age determination, and reduce the error of reading.

Key words: Artificial intelligence software, Developmental related diseases in children, Bone age, Reading time, Accuracy, Error

中图分类号:

R179

邓成清, 孙安林, 唐宇轩, 张萍, 唐爱群. 人工智能软件在儿童发育相关疾病骨龄评估中的临床应用[J]. 中华养生保健, 2023, 41(22): 54-57.

DENG Cheng-qing, SUN An-lin, TANG Yu-xuan, ZHANG Ping, TANG Ai-qun. Clinical Application of Artificial Intelligence Software in Bone Age Assessment of Developmental Related Diseases in Children[J]. ZHONGHUA YANGSHENG BAOJIAN, 2023, 41(22): 54-57.

参考文献

[1] 文颖,任旭华,杨秀军,等.基于手腕部影像传统关注特征区域深度学习的人工智能骨龄评估[J].中华放射学杂志,2019,53(10):895-899.
[2] 次旦旺久,土旦阿旺,杨美杰,等.海拔高度对儿童及青少年骨龄发育的影响[J].基础医学与临床,2023,43(4):636-640.
[3] 袁新宇,潘诗农,程晓光.构建儿童骨龄核心数据,建立中国儿童骨龄标准体系[J].中华放射学杂志,2023,57(4):341-343.
[4] 常沙,闫东,杜霞,等.基于深度学习的骨龄评估系统对生长发育异常儿童骨龄评测的准确性[J].中华放射学杂志,2023,57(4):364-369.
[5] 麻宏伟,李琳.儿童发育障碍相关疾病的病因鉴别和诊断[J].中国实用儿科杂志,2016,31(10):751-756.
[6] 张义,朱文文,李凯,等.人工智能在儿童骨龄影像评估的应用进展[J].中华放射学杂志,2023,57(4):420-423.
[7] 许可,宁刚.基于人工智能的骨龄辅助评价系统对四川地区完全性生长激素缺乏症患儿骨龄研究[J].中华妇幼临床医学杂志(电子版),2022,18(4):449-459.
[8] 钱明珠,吴澄,徐芳,等.矮小儿童的诊断和治疗[J].中国育生优育,2001,12(2):53-56.
[9] 王岩,霍爱华,王大为,等.基于深度学习人工智能骨龄测评系统临床应用[J].中国医学影像技术,2021,37(1):104-107.
[10] 赵凯,马帅,孙佳丽,等.人工智能软件对住院医师X线骨龄诊断辅助效果初探[J].实用放射学杂志,2021,37(2):317-320.
[11] 占梦军,张世杰,刘力,等.基于深度学习自动化评估四川汉族青少年左手腕关节骨龄[J].中国法医学杂志,2019,34(5):427-432.
[12] 彭乐媛,段玉梅,王永芹,等.4种新型人工智能算法在区域人群骨龄测定中的价值[J].中国医学装备,2021,18(11):113-117.
[13] 李莉红,杨秀军,李婷婷.不同人工智能模型对基于手腕部DR影像的骨龄预测比较[J].临床放射学杂志,2019,38(8):1498-1501.
[14] 白凤森,袁新宇,马毅民,等.人工智能系统对于儿童骨龄评价效率提升的研究[J].医学影像学杂志,2021,31(12):2111-2113,2140.
[15] 李然然,杨芸晓,高剑波,等.人工智能支持系统辅助放射科医师评估骨龄[J].临床放射学杂志,2021,40(3):560-563.
[16] 林紫薇,吴山东,曲伸.人工智能在内分泌代谢领域的应用及展望[J].中华内分泌代谢杂志,2021,37(2):89-93.
[17] 次旦旺久,拉巴顿珠,王凤丹,等.三种方法评估藏族儿童骨龄效果比较及藏族儿童骨龄发育特点[J].协和医学杂志,2021,12(3):411-416.
[18] 施莉丽,杨秀军,于广军,等.基于医院RIS-PACS场景的人工智能骨龄检测系统集成技术与实现[J].中国医疗器械杂志,2020,44(5):415-419.
[19] 黄卫保,林剑军,梁莎,等.国内外手腕部骨龄影像评估方法各自优缺点及研究进展[J].中国医疗设备,2020,35(10):181-185.
[20] 宋娟,宫平,高畅,等.基于深度学习的儿童骨龄智能评估模型构建及初步临床验证[J].中华放射学杂志,2019,53(11):974-978.