Author:
Jiang Yaoxian,Yang Guangyao,Liang Yuan,Shi Qin,Cui Boqi,Chang Xiaodan,Qiu Zhaowen,Zhao Xudong
Abstract
PurposeA computer-aided system was used to semiautomatically measure Tönnis angle, Sharp angle, and center-edge (CE) angle using contours of the hip bones to establish an auxiliary measurement model for developmental screening or diagnosis of hip joint disorders.MethodsWe retrospectively analyzed bilateral hip x-rays for 124 patients (41 men and 83 women aged 20–70 years) who presented at the Affiliated Zhongshan Hospital of Dalian University in 2017 and 2018. All images were imported into a computer-aided detection system. After manually outlining hip bone contours, Tönnis angle, Sharp angle, and CE angle marker lines were automatically extracted, and the angles were measured and recorded. An imaging physician also manually measured all angles and recorded hip development, and Pearson correlation coefficients were used to compare computer-aided system measurements with imaging physician measurements. Accuracy for different angles was calculated, and the area under the receiver operating characteristic (AUROC) curve was used to represent the diagnostic efficiency of the computer-aided system.ResultsFor Tönnis angle, Sharp angle, and CE angle, correlation coefficients were 0.902, 0.887, and 0.902, respectively; the accuracies of the computer-aided detection system were 89.1, 93.1, and 82.3%; and the AUROC curve values were 0.940, 0.956, and 0.948.ConclusionThe measurements of Tönnis angle, Sharp angle, and CE angle using the semiautomatic system were highly correlated with the measurements of the imaging physician and can be used to assess hip joint development with high accuracy and diagnostic efficiency.
Subject
Physiology (medical),Physiology
Reference37 articles.
1. Imaging evaluation of developmental hip dysplasia in the young adult.;Beltran;AJR Am. J. Roentgenol.,2013
2. Artificial intelligence in cancer imaging: clinical challenges and applications.;Bi;CA Cancer J. Clin.,2019
3. Current applications and future impact of machine learning in radiology.;Choy;Radiology,2018
4. Acetabular dysplasia in the adult.;Cooperman;Clin. Orthop. Relat. Res.,1983
5. Small pulmonary nodules: effect of two computer-aided detection systems on radiologist performance.;Das;Radiology,2006
Cited by
4 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献