基于空间结构和注意机制的卷积神经网络taner - whitehouse 3骨龄评估

IF 4.9 2区医学 Q1 ENGINEERING, BIOMEDICAL

Biomedical Signal Processing and Control Pub Date : 2025-10-04 DOI:10.1016/j.bspc.2025.108728

Yi Zhang , Jindong Wu , Wenshuang Zhang , Hongye Zhao , Kai Li , Jian Geng , Dong Yan , Xiaoguang Cheng , Tongning Wu

{"title":"基于空间结构和注意机制的卷积神经网络taner - whitehouse 3骨龄评估","authors":"Yi Zhang , Jindong Wu , Wenshuang Zhang , Hongye Zhao , Kai Li , Jian Geng , Dong Yan , Xiaoguang Cheng , Tongning Wu","doi":"10.1016/j.bspc.2025.108728","DOIUrl":null,"url":null,"abstract":"<div><h3>Objective</h3><div>Radiographic bone age assessment (BAA) is a standard clinical procedure for the diagnosis of skeletal growth abnormalities in children and infants. Existing automated BAA algorithms based on the Tanner-Whitehouse 3 (TW3) method can only assess the skeletal maturity and bone age of the radius, ulna, and short bone (TW3-RUS), lacking the capability to assess the bone age of the carpal bone (TW3-C), which hinders their wider clinical adoption.</div></div><div><h3>Methods</h3><div>We proposed a TW3-based automated BAA method to address this limitation. Firstly, a heat map regression key point detection algorithm incorporating spatial configurations was introduced to locate and segment all 20 TW3-regions of interest (ROIs). Subsequently, a skeletal maturity classification network incorporating an attention mechanism with spatial and channel features was proposed to predict the skeletal maturity scores and bone ages of the TW3-RUS and TW3-C series.</div></div><div><h3>Results</h3><div>Our approach achieved a mean absolute error (MAE) of bone age of 0.42 (TW3-RUS) and 0.44 (TW3-C) years on a dataset of 5,235 left lateral radiographs of children of different ages.</div></div><div><h3>Conclusions</h3><div>Our framework demonstrated the immense clinical potential of the proposed algorithm by achieving the impressive BAA results, while also providing clinicians with all the essential information they need to know about skeletal maturity and bone age.</div></div><div><h3>Significance</h3><div>The proposed BAA algorithm which can simultaneously evaluate skeletal maturity level and bone age for both the TW3-RUS and TW3-C series is more helpful for clinicians to analyze the progression of their patients’ conditions and to adjust their treatment plans.</div></div>","PeriodicalId":55362,"journal":{"name":"Biomedical Signal Processing and Control","volume":"112 ","pages":"Article 108728"},"PeriodicalIF":4.9000,"publicationDate":"2025-10-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"A convolutional neural networks with spatial configuration and attention mechanism for Tanner-Whitehouse 3 bone age assessment\",\"authors\":\"Yi Zhang , Jindong Wu , Wenshuang Zhang , Hongye Zhao , Kai Li , Jian Geng , Dong Yan , Xiaoguang Cheng , Tongning Wu\",\"doi\":\"10.1016/j.bspc.2025.108728\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><h3>Objective</h3><div>Radiographic bone age assessment (BAA) is a standard clinical procedure for the diagnosis of skeletal growth abnormalities in children and infants. Existing automated BAA algorithms based on the Tanner-Whitehouse 3 (TW3) method can only assess the skeletal maturity and bone age of the radius, ulna, and short bone (TW3-RUS), lacking the capability to assess the bone age of the carpal bone (TW3-C), which hinders their wider clinical adoption.</div></div><div><h3>Methods</h3><div>We proposed a TW3-based automated BAA method to address this limitation. Firstly, a heat map regression key point detection algorithm incorporating spatial configurations was introduced to locate and segment all 20 TW3-regions of interest (ROIs). Subsequently, a skeletal maturity classification network incorporating an attention mechanism with spatial and channel features was proposed to predict the skeletal maturity scores and bone ages of the TW3-RUS and TW3-C series.</div></div><div><h3>Results</h3><div>Our approach achieved a mean absolute error (MAE) of bone age of 0.42 (TW3-RUS) and 0.44 (TW3-C) years on a dataset of 5,235 left lateral radiographs of children of different ages.</div></div><div><h3>Conclusions</h3><div>Our framework demonstrated the immense clinical potential of the proposed algorithm by achieving the impressive BAA results, while also providing clinicians with all the essential information they need to know about skeletal maturity and bone age.</div></div><div><h3>Significance</h3><div>The proposed BAA algorithm which can simultaneously evaluate skeletal maturity level and bone age for both the TW3-RUS and TW3-C series is more helpful for clinicians to analyze the progression of their patients’ conditions and to adjust their treatment plans.</div></div>\",\"PeriodicalId\":55362,\"journal\":{\"name\":\"Biomedical Signal Processing and Control\",\"volume\":\"112 \",\"pages\":\"Article 108728\"},\"PeriodicalIF\":4.9000,\"publicationDate\":\"2025-10-04\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Biomedical Signal Processing and Control\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S174680942501239X\",\"RegionNum\":2,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"ENGINEERING, BIOMEDICAL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Biomedical Signal Processing and Control","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S174680942501239X","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, BIOMEDICAL","Score":null,"Total":0}

引用次数: 0

摘要

目的x线骨龄评估（BAA）是诊断儿童和婴儿骨骼生长异常的标准临床程序。现有基于Tanner-Whitehouse 3 （TW3）方法的自动BAA算法仅能评估桡骨、尺骨、短骨（TW3- rus）的骨骼成熟度和骨龄，缺乏评估腕骨（TW3- c）骨龄的能力，阻碍了其在临床的广泛应用。方法提出了一种基于tw3的自动化BAA方法来解决这一问题。首先，引入融合空间构型的热图回归关键点检测算法，对20个tw3感兴趣区域进行定位和分割；随后，我们提出了一个包含空间和通道特征的注意机制的骨骼成熟度分类网络，用于预测TW3-RUS和TW3-C系列的骨骼成熟度评分和骨年龄。结果我们的方法获得的骨龄平均绝对误差（MAE）为0.42 （TW3-RUS）和0.44 （TW3-C）年，数据集为5,235名不同年龄的儿童左侧x线片。通过获得令人印象深刻的BAA结果，sour框架展示了所提出算法的巨大临床潜力，同时也为临床医生提供了他们需要了解的骨骼成熟度和骨龄的所有基本信息。意义提出的BAA算法可以同时评估TW3-RUS和TW3-C系列的骨骼成熟度水平和骨龄，更有助于临床医生分析患者病情的进展，调整治疗方案。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

A convolutional neural networks with spatial configuration and attention mechanism for Tanner-Whitehouse 3 bone age assessment

Objective

Radiographic bone age assessment (BAA) is a standard clinical procedure for the diagnosis of skeletal growth abnormalities in children and infants. Existing automated BAA algorithms based on the Tanner-Whitehouse 3 (TW3) method can only assess the skeletal maturity and bone age of the radius, ulna, and short bone (TW3-RUS), lacking the capability to assess the bone age of the carpal bone (TW3-C), which hinders their wider clinical adoption.

Methods

We proposed a TW3-based automated BAA method to address this limitation. Firstly, a heat map regression key point detection algorithm incorporating spatial configurations was introduced to locate and segment all 20 TW3-regions of interest (ROIs). Subsequently, a skeletal maturity classification network incorporating an attention mechanism with spatial and channel features was proposed to predict the skeletal maturity scores and bone ages of the TW3-RUS and TW3-C series.

Results

Our approach achieved a mean absolute error (MAE) of bone age of 0.42 (TW3-RUS) and 0.44 (TW3-C) years on a dataset of 5,235 left lateral radiographs of children of different ages.

Conclusions

Our framework demonstrated the immense clinical potential of the proposed algorithm by achieving the impressive BAA results, while also providing clinicians with all the essential information they need to know about skeletal maturity and bone age.

Significance

The proposed BAA algorithm which can simultaneously evaluate skeletal maturity level and bone age for both the TW3-RUS and TW3-C series is more helpful for clinicians to analyze the progression of their patients’ conditions and to adjust their treatment plans.

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

Biomedical Signal Processing and Control 工程技术-工程：生物医学

CiteScore

9.80

自引率

13.70%

发文量

822

审稿时长

4 months

期刊介绍： Biomedical Signal Processing and Control aims to provide a cross-disciplinary international forum for the interchange of information on research in the measurement and analysis of signals and images in clinical medicine and the biological sciences. Emphasis is placed on contributions dealing with the practical, applications-led research on the use of methods and devices in clinical diagnosis, patient monitoring and management. Biomedical Signal Processing and Control reflects the main areas in which these methods are being used and developed at the interface of both engineering and clinical science. The scope of the journal is defined to include relevant review papers, technical notes, short communications and letters. Tutorial papers and special issues will also be published.