Literature DB >> 32269926

Diagnostic performance of convolutional neural network-based Tanner-Whitehouse 3 bone age assessment system.

Xue-Lian Zhou1, Er-Gang Wang2,3, Qiang Lin4, Guan-Ping Dong1, Wei Wu1, Ke Huang1, Can Lai5, Gang Yu6, Hai-Chun Zhou5, Xiao-Hui Ma5, Xuan Jia5, Lei Shi4, Yong-Sheng Zheng4, Lan-Xuan Liu4, Da Ha4, Hao Ni4, Jun Yang4, Jun-Fen Fu1.   

Abstract

BACKGROUND: Bone age can reflect the true growth and development status of a child; thus, it plays a critical role in evaluating growth and endocrine disorders. This study established and validated an optimized Tanner-Whitehouse 3 artificial intelligence (TW3-AI) bone age assessment (BAA) system based on a convolutional neural network (CNN).
METHODS: A data set of 9,059 clinical radiographs of the left hand was obtained from the picture archives and communication systems (PACS) between January 2012 and December 2016. Among these, 8,005/9,059 (88%) samples were treated as the training set for model implementation, 804/9,059 (9%) samples as the validation set for parameters optimization, and the remaining 250/9,059 (3%) samples were used to verify the accuracy and reliability of the model compared to that of 4 experienced endocrinologists and 2 experienced radiologists. The overall variation of TW3-metacarpophalangeal, radius, ulna and short bones (RUS) and TW3-Carpal bone score, as well as each bone (13 RUS + 7 Carpal) between reviewers and the AI, were compared by Bland-Altman (BA) chart and Kappa test, respectively. Furthermore, the time consumption between the model and reviewers was also compared.
RESULTS: The performance of TW3-AI model was highly consistent with the reviewers' overall estimation, and the root mean square (RMS) was 0.50 years. The accuracy of the BAA of the TW3-AI model was better than the estimate of the reviewers. Further analysis revealed that human interpretations of the male capitate, hamate, the first distal and fifth middle phalanx and female capitate, the trapezoid, and the third and fifth middle phalanx, were most inconsistent. The average image processing time was 1.5±0.2 s in the TW3-AI model, which was significantly shorter than manual interpretation.
CONCLUSIONS: The diagnostic performance of CNN-based TW3 BAA was accurate and timesaving, and possesses better stability compared to diagnostics made by experienced experts. 2020 Quantitative Imaging in Medicine and Surgery. All rights reserved.

Entities:  

Keywords:  Artificial intelligence (AI); Tanner-Whitehouse 3 method (TW3 method); bone age; convolutional neural network (CNN)

Year:  2020        PMID: 32269926      PMCID: PMC7136746          DOI: 10.21037/qims.2020.02.20

Source DB:  PubMed          Journal:  Quant Imaging Med Surg        ISSN: 2223-4306


  19 in total

1.  Bone age assessment: a large scale comparison of the Greulich and Pyle, and Tanner and Whitehouse (TW2) methods.

Authors:  R K Bull; P D Edwards; P M Kemp; S Fry; I A Hughes
Journal:  Arch Dis Child       Date:  1999-08       Impact factor: 3.791

2.  Computer-assisted phalangeal analysis in skeletal age assessment.

Authors:  E Pietka; M F McNitt-Gray; M L Kuo; H K Huang
Journal:  IEEE Trans Med Imaging       Date:  1991       Impact factor: 10.048

Review 3.  Bone age assessment: automated techniques coming of age?

Authors:  R R van Rijn; H H Thodberg
Journal:  Acta Radiol       Date:  2013-11       Impact factor: 1.990

4.  Bone age assessment in young children using automatic carpal bone feature extraction and support vector regression.

Authors:  Krit Somkantha; Nipon Theera-Umpon; Sansanee Auephanwiriyakul
Journal:  J Digit Imaging       Date:  2011-12       Impact factor: 4.056

Review 5.  Are the new automated methods for bone age estimation advantageous over the manual approaches?

Authors:  Vincenzo De Sanctis; Ashraf T Soliman; Salvatore Di Maio; Said Bedair
Journal:  Pediatr Endocrinol Rev       Date:  2014-12

6.  Effect of training on replicability of assessments of skeletal maturity (Greulich-Pyle).

Authors:  A F Roche; C G Rohmann; N Y French; G H Dávila
Journal:  Am J Roentgenol Radium Ther Nucl Med       Date:  1970-03

7.  Performance of a Deep-Learning Neural Network Model in Assessing Skeletal Maturity on Pediatric Hand Radiographs.

Authors:  David B Larson; Matthew C Chen; Matthew P Lungren; Safwan S Halabi; Nicholas V Stence; Curtis P Langlotz
Journal:  Radiology       Date:  2017-11-02       Impact factor: 11.105

8.  Dermatologist-level classification of skin cancer with deep neural networks.

Authors:  Andre Esteva; Brett Kuprel; Roberto A Novoa; Justin Ko; Susan M Swetter; Helen M Blau; Sebastian Thrun
Journal:  Nature       Date:  2017-01-25       Impact factor: 49.962

9.  3D convolutional neural network for differentiating pre-invasive lesions from invasive adenocarcinomas appearing as ground-glass nodules with diameters ≤3 cm using HRCT.

Authors:  Shengping Wang; Rui Wang; Shengjian Zhang; Ruimin Li; Yi Fu; Xiangjie Sun; Yuan Li; Xing Sun; Xinyang Jiang; Xiaowei Guo; Xuan Zhou; Jia Chang; Weijun Peng
Journal:  Quant Imaging Med Surg       Date:  2018-06

10.  The RSNA Pediatric Bone Age Machine Learning Challenge.

Authors:  Safwan S Halabi; Luciano M Prevedello; Jayashree Kalpathy-Cramer; Artem B Mamonov; Alexander Bilbily; Mark Cicero; Ian Pan; Lucas Araújo Pereira; Rafael Teixeira Sousa; Nitamar Abdala; Felipe Campos Kitamura; Hans H Thodberg; Leon Chen; George Shih; Katherine Andriole; Marc D Kohli; Bradley J Erickson; Adam E Flanders
Journal:  Radiology       Date:  2018-11-27       Impact factor: 29.146

View more
  7 in total

1.  SMANet: multi-region ensemble of convolutional neural network model for skeletal maturity assessment.

Authors:  Yi Zhang; Wenwen Zhu; Kai Li; Dong Yan; Hua Liu; Jie Bai; Fan Liu; Xiaoguang Cheng; Tongning Wu
Journal:  Quant Imaging Med Surg       Date:  2022-07

2.  A Randomized Controlled Phase 3 Study on the Efficacy and Safety of Recombinant Human Growth Hormone in Children With Idiopathic Short Stature.

Authors:  Jinna Yuan; Junfen Fu; Haiyan Wei; Gaixiu Zhang; Yanfeng Xiao; Hongwei Du; Wei Gu; Yanhong Li; Linqi Chen; Feihong Luo; Yan Zhong; Haihong Gong
Journal:  Front Endocrinol (Lausanne)       Date:  2022-04-29       Impact factor: 6.055

3.  Data governance system of the National Clinical Research Center for Child Health in China.

Authors:  Jing Li; Gang Yu; Wen Ding; Jian Huang; Zheming Li; Zhu Zhu; Dejian Wang; Jie Zhang; Jing Wang; Jianwei Yin
Journal:  Transl Pediatr       Date:  2021-07

4.  Artificial Intelligence-Assisted Bone Age Assessment to Improve the Accuracy and Consistency of Physicians With Different Levels of Experience.

Authors:  Xi Wang; Bo Zhou; Ping Gong; Ting Zhang; Yan Mo; Jie Tang; Xinmiao Shi; Jianhong Wang; Xinyu Yuan; Fengsen Bai; Lei Wang; Qi Xu; Yu Tian; Qing Ha; Chencui Huang; Yizhou Yu; Lin Wang
Journal:  Front Pediatr       Date:  2022-02-24       Impact factor: 3.418

5.  Association between phthalate exposure and accelerated bone maturation in Chinese girls with early puberty onset: a propensity score-matched case-control analysis.

Authors:  Shurong Huang; Zhe Su; Huiping Su; Yanhua Jiao; Qiru Su; Yao Yao; Li Zhou; Xiuxin Zheng; Xingliang Zhang
Journal:  Sci Rep       Date:  2022-09-07       Impact factor: 4.996

6.  A comparative study of three bone age assessment methods on Chinese preschool-aged children.

Authors:  Chengcheng Gao; Qi Qian; Yangsheng Li; Xiaowei Xing; Xiao He; Min Lin; Zhongxiang Ding
Journal:  Front Pediatr       Date:  2022-08-16       Impact factor: 3.569

7.  Performance of an artificial intelligence system for bone age assessment in Tibet.

Authors:  Fengdan Wang; Wangjiu Cidan; Xiao Gu; Shi Chen; Wu Yin; Yongliang Liu; Lei Shi; Hui Pan; Zhengyu Jin
Journal:  Br J Radiol       Date:  2021-02-09       Impact factor: 3.039

  7 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.