Bone Age Prediction Based on Lightweight VGG16 and Attention Mechanism

GUO Wei; GUO Jin-xing; CHEN Guang-xin; HAN Xue-shan

doi:10.3969/j.issn.2096-6091.2023.19.003

您当前的位置：

首页 >

文章列表页 >

Bone Age Prediction Based on Lightweight VGG16 and Attention Mechanism

RESEARCH PAPER | 更新时间：2025-03-26

- Bone Age Prediction Based on Lightweight VGG16 and Attention Mechanism
- New Generation of Information Technology Vol. 6, Issue 19, Pages: 15-20(2023)
- 作者机构：
  
  1.牡丹江医学院生命科学学院，黑龙江牡丹江 157011
  2.牡丹江医学院附属红旗医院，黑龙江牡丹江 157011
  3.牡丹江医学院医学影像学院，黑龙江牡丹江 157011
- 作者简介：
- 基金信息：
- DOI：10.3969/j.issn.2096-6091.2023.19.003
  CLC： TP30
- Published：15 October 2023
- 稿件说明：
移动端阅览
国威, 郭金兴, 陈广新, 等. 基于轻量化VGG16和注意力机制的骨龄预测研究[J]. 新一代信息技术, 2023, 6(19): 15-20

GUO Wei, GUO Jin-xing, CHEN Guang-xin, et al. Bone Age Prediction Based on Lightweight VGG16 and Attention Mechanism[J]. New Generation of Information Technology, 2023, 6(19): 15-20
国威, 郭金兴, 陈广新, 等. 基于轻量化VGG16和注意力机制的骨龄预测研究[J]. 新一代信息技术, 2023, 6(19): 15-20 DOI： 10.3969/j.issn.2096-6091.2023.19.003.

GUO Wei, GUO Jin-xing, CHEN Guang-xin, et al. Bone Age Prediction Based on Lightweight VGG16 and Attention Mechanism[J]. New Generation of Information Technology, 2023, 6(19): 15-20 DOI： 10.3969/j.issn.2096-6091.2023.19.003.

摘要

本研究旨在提高儿童骨龄预测的准确性，通过引入轻量化VGG16和CBAM（Convolutional Block Attention Module）注意力机制，我们设计了轻量化VGG16网络，通过迁移学习初始化模型，同时利用CBAM注意力机制提取关键特征。在儿童骨龄预测数据集上，轻量化VGG16网络相较于传统VGG16、ResNet34和MobileNetV2，表现出更高的预测性能，MAE（Mean Absolute Error）为5.01（月）。实验结果及图像对比验证了模型的优越性，为提高儿童骨龄预测提供了可靠的支持。

Abstract

This study aims to improve the accuracy of bone age prediction in children. By introducing lightweight VGG16 and CBAM (Convolutional Block Attention Module) attention mechanisms

we designed a lightweight VGG16 network

initialized the model through transfer learning

and extracted key features using CBAM attention mechanisms. Compared with traditional VGG16

ResNet34 and MobileNetV2

the lightweight VGG16 network showed higher predictive performance in the child bone age prediction dataset

with a MAE (Mean Absolute Error) of 5.01 (month). The experimental results and image comparison verified the superiority of the model

and provided reliable support for improving the prediction of children's bone age.

关键词

Keywords

references

CREO A L , 2nd SCHWENK W F . Bone age: A handy tool for pediatric providers [J ] . Pediatrics , 2017 , 140 ( 6 ): e20171486 .

HALABI S S , PREVEDELLO L M , KALPATHY-CRAMER J , et al . The RSNA pediatric bone age machine learning challenge [J ] . Radiology , 2019 , 290 ( 2 ): 498 - 503 .

OSTOJIC S M . Prediction of adult height by Tanner-Whitehouse method in young Caucasian male athletes [J ] . QJM , 2013 , 106 ( 4 ): 341 - 345 .

TAJMIR S H , LEE H , SHAILAM R , et al . Artificial intelligence-assisted interpretation of bone age radiographs improves accuracy and decreases variability [J ] . Skeletal Radiology , 2019 , 48 ( 2 ): 275 - 283 .

ACHESON R M , VICINUS J H , FOWLER G B . Studies in the reliability of assessing skeletal maturity from X-rays. 3. Greulich-Pyle Atlas and Tanner-Whitehouse method contrasted [J ] . Human Biology , 1966 , 38 ( 3 ): 204 - 218 .

TEELE R L . Assessment of skeletal maturity and prediction of adult height [J ] . Journal of Paediatrics and Child Health , 2003 , 39 ( 4 ): 322 - 323 .

张绍岩 , 刘丽娟 , 吴真列 , 等 . 中国人手腕骨发育标准-中华05 I.TW3-C RUS、TW3-C腕骨和RUS-CHN方法 [J ] . 中国运动医学杂志 , 2006 , 25 ( 5 ): 509 - 516 .

LARSON D B , CHEN M C , LUNGREN M P , et al . Performance of a deep-learning neural network model in assessing skeletal maturity on pediatric hand radiographs [J ] . Radiology , 2018 , 287 ( 1 ): 313 - 322 .

SALIM I , BEN HAMZA A . Ridge regression neural network for pediatric bone age assessment [J ] . Multimedia Tools and Applications , 2021 , 80 ( 20 ): 30461 - 30478 .

LEE H , TAJMIR S , LEE J , et al . Fully automated deep learning system for bone age assessment [J ] . Journal of Digital Imaging , 2017 , 30 ( 4 ): 427 - 441 .

谷静 , 孙启雷 , 张明选 . 基于TW3-C RUS法的骨龄评估方法研究 [J ] . 光电子•激光 , 2023 , 34 ( 5 ): 554 - 560 .

BEHESHTIAN E , PUTMAN S K . SANTOMARTINO M, 等. 基于手部X线片的深度学习儿童骨龄预测模型的通用性和偏差[J ] . 国际医学放射学杂志 , 2023 , 46 ( 2 ): e220505 .

郭子昇 , 王吉芳 , 苏鹏 . 基于深度学习的智能骨龄分类器 [J ] . 计算机系统应用 , 2022 , 31 ( 6 ): 339 - 346 .

BRAHIMI S , BEN AOUN N , BEN AMAR C . Boosted Convolutional Neural Network for object recognition at large scale [J ] . Neurocomputing , 2019 , 330 : 337 - 354 .

WANG S H , FERNANDES S L , ZHU Z , et al . AVNC: Attention-based VGG-style network for COVID-19 diagnosis by CBAM [J ] . IEEE Sensors Journal , 2022 , 22 ( 18 ): 17431 - 17438 .

Views

250

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Study on the Recognition of Maize Leaf Disease Based on Deep Learning with Small Sample Size

Predictive Study of Breast Cancer Pathological Images Based on DenseNet201

Application and Prospect of Deep Learning Technology in Medical Image Analysis

Related Author

HAN Xueshan

Chen Guangxin

GUO Jinxing

LI Ting-ting

WANG Qing-qing

TANG Qi

ZHANG Hao

HUI Xiang-hui

Related Institution

Henan Agricultural University ， College of Information and Mangement Science

Hongqi Hospital Affiliated with Mudanjiang Medical University

Basic Medical College of Mudanjiang Medical University

Medical Imaging College of Mudanjiang Medical University

Chongqing Communication Design Institute Co., Ltd

⁰