TW3-Based Fully Automated Bone Age Assessment System Using Deep Neural Networks
Deep learning technology has rapidly evolved in recent years. Bone age assessment (BAA) is a typical object detection and classification problem that would benefit from deep learning. Convolutional neural networks (CNNs) and their variants are hence increasingly used for automating BAA, and they hav...
Main Authors: | , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2019-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/8660640/ |
id |
doaj-4d111a6d64544b398518587c143331c5 |
---|---|
record_format |
Article |
spelling |
doaj-4d111a6d64544b398518587c143331c52021-03-29T22:58:24ZengIEEEIEEE Access2169-35362019-01-017333463335810.1109/ACCESS.2019.29031318660640TW3-Based Fully Automated Bone Age Assessment System Using Deep Neural NetworksSung Joon Son0Youngmin Song1Namgi Kim2Younghae Do3Nojun Kwak4Mu Sook Lee5Byoung-Dai Lee6https://orcid.org/0000-0002-4028-6168Graduate School of Convergence Science and Technology, Seoul National University, Seoul, South KoreaDepartment of Computer Science, Kyonngi University, Suwon, South KoreaDepartment of Computer Science, Kyonngi University, Suwon, South KoreaDepartment of Mathematics, Kyungpook National University, Daegu, South KoreaGraduate School of Convergence Science and Technology, Seoul National University, Seoul, South KoreaHuman Medical Imaging Center, Seoul, South KoreaDepartment of Computer Science, Kyonngi University, Suwon, South KoreaDeep learning technology has rapidly evolved in recent years. Bone age assessment (BAA) is a typical object detection and classification problem that would benefit from deep learning. Convolutional neural networks (CNNs) and their variants are hence increasingly used for automating BAA, and they have shown promising results. In this paper, we propose a complete end-to-end BAA system to automate the entire process of the Tanner-Whitehouse 3 method, starting from localization of the epiphysis-metaphysis growth regions within 13 different bones and ending with estimation of the corresponding BA. Specific modifications to the CNNs and other stages are proposed to improve results. In addition, an annotated database of 3300 X-ray images is built to train and evaluate the system. The experimental results show that the average top-1 and top-2 prediction accuracies for skeletal bone maturity levels for 13 regions of interest are 79.6% and 97.2%, respectively. The mean absolute error and root mean squared error in age prediction are 0.46 years and 0.62 years, respectively, and accuracy within one year of the ground truth of 97.6% is achieved. The proposed system is shown to outperform a commercially available Greulich-Pyle-based system, demonstrating the potential for practical clinical use.https://ieeexplore.ieee.org/document/8660640/Bone age assessmentdeep learningGPTW3 |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Sung Joon Son Youngmin Song Namgi Kim Younghae Do Nojun Kwak Mu Sook Lee Byoung-Dai Lee |
spellingShingle |
Sung Joon Son Youngmin Song Namgi Kim Younghae Do Nojun Kwak Mu Sook Lee Byoung-Dai Lee TW3-Based Fully Automated Bone Age Assessment System Using Deep Neural Networks IEEE Access Bone age assessment deep learning GP TW3 |
author_facet |
Sung Joon Son Youngmin Song Namgi Kim Younghae Do Nojun Kwak Mu Sook Lee Byoung-Dai Lee |
author_sort |
Sung Joon Son |
title |
TW3-Based Fully Automated Bone Age Assessment System Using Deep Neural Networks |
title_short |
TW3-Based Fully Automated Bone Age Assessment System Using Deep Neural Networks |
title_full |
TW3-Based Fully Automated Bone Age Assessment System Using Deep Neural Networks |
title_fullStr |
TW3-Based Fully Automated Bone Age Assessment System Using Deep Neural Networks |
title_full_unstemmed |
TW3-Based Fully Automated Bone Age Assessment System Using Deep Neural Networks |
title_sort |
tw3-based fully automated bone age assessment system using deep neural networks |
publisher |
IEEE |
series |
IEEE Access |
issn |
2169-3536 |
publishDate |
2019-01-01 |
description |
Deep learning technology has rapidly evolved in recent years. Bone age assessment (BAA) is a typical object detection and classification problem that would benefit from deep learning. Convolutional neural networks (CNNs) and their variants are hence increasingly used for automating BAA, and they have shown promising results. In this paper, we propose a complete end-to-end BAA system to automate the entire process of the Tanner-Whitehouse 3 method, starting from localization of the epiphysis-metaphysis growth regions within 13 different bones and ending with estimation of the corresponding BA. Specific modifications to the CNNs and other stages are proposed to improve results. In addition, an annotated database of 3300 X-ray images is built to train and evaluate the system. The experimental results show that the average top-1 and top-2 prediction accuracies for skeletal bone maturity levels for 13 regions of interest are 79.6% and 97.2%, respectively. The mean absolute error and root mean squared error in age prediction are 0.46 years and 0.62 years, respectively, and accuracy within one year of the ground truth of 97.6% is achieved. The proposed system is shown to outperform a commercially available Greulich-Pyle-based system, demonstrating the potential for practical clinical use. |
topic |
Bone age assessment deep learning GP TW3 |
url |
https://ieeexplore.ieee.org/document/8660640/ |
work_keys_str_mv |
AT sungjoonson tw3basedfullyautomatedboneageassessmentsystemusingdeepneuralnetworks AT youngminsong tw3basedfullyautomatedboneageassessmentsystemusingdeepneuralnetworks AT namgikim tw3basedfullyautomatedboneageassessmentsystemusingdeepneuralnetworks AT younghaedo tw3basedfullyautomatedboneageassessmentsystemusingdeepneuralnetworks AT nojunkwak tw3basedfullyautomatedboneageassessmentsystemusingdeepneuralnetworks AT musooklee tw3basedfullyautomatedboneageassessmentsystemusingdeepneuralnetworks AT byoungdailee tw3basedfullyautomatedboneageassessmentsystemusingdeepneuralnetworks |
_version_ |
1724190424617189376 |