QR Code
eISSN 2005-372X
QR Code
All Articles
View
Split Viewer
Korean J Orthod
First Published Date October 18, 2023
Copyright © The Korean Association of Orthodontists.
1st author: Sung-Hoon Han,a, 1st author: Jisup Lim,c, Jun-Sik Kim,d, Jin-Hyoung Cho,e, Mihee Hong,f, Minji Kim,g, Su-Jung Kim,h, Yoon-Ji Kim,i, Young Ho Kim,j, Sung-Hoon Lim,k, Sang Jin Sung,i, Kyung-Hwa Kang,a, Seung-Hak Baek,l, corresponding author: Sung-Kwon Choi,a, corresponding author: Namkug Kim,b
aDepartment of Orthodontics, School of Dentistry, Wonkwang University, Iksan-si, South Korea.
bDepartment of Convergence Medicine, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea.
cDepartment of Convergence Medicine, University of Ulsan, College of Medicine, Asan Medical Center, Seoul, Republic of Korea
dGraduate student (MSc), Department of Biomedical Engineering, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea
eDepartment of Orthodontics, School of Dentistry, Chonnam National University, Gwangju, South Korea.
fDepartment of Orthodontics, School of Dentistry, Kyungpook National University, Daegu, South Korea.
gDepartment of Orthodontics, College of Medicine, Ewha Womans University, Seoul, South Korea.
hDepartment of Orthodontics, School of Dentistry, Kyung Hee University, Seoul, South Korea.
iDepartment of Orthodontics, Asan Medical Center, College of Medicine, University of Ulsan, Seoul, South Korea.
jDepartment of Orthodontics, Institute of Oral Health Science, Ajou University School of Medicine, Suwon-si, South Korea.
kDepartment of Orthodontics, College of Dentistry, Chosun University, Gwangju, South Korea.
lDepartment of Orthodontics, School of Dentistry, Dental Research Institute, Seoul National University, Seoul, South Korea.
Correspondence to:Sung-Kwon Choi, Department of Orthodontics, School of Dentistry, Wonkwang University, 460 Iksandae-ro, Iksan-si, Jeollabuk-do 54538, South Korea; Tel: +82-63-859-2962; e-mail, chsk6206@wku.ac.kr
or
Namkug Kim, Department of Convergence Medicine, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, 88 Olympic-Ro 43-Gil, Songpa-Gu, Seoul 05505, South Korea; Tel: +82-2-3010-6573; e-mail, namkugkim@gmail.com.
Objective: To quantify the effects of midline-related landmark identification in posteroanterior (PA) cephalogram images by cascaded convolutional neural network (CNN) algorithm on the midline deviation measurements.
Methods: A total of 2,903 PA cephalogram images obtained from nine university hospitals were divided into the training-set (n=2,150), internal validation-set (n=376), and test-set (n=377). As gold standard, two orthodontic professors marked the bilateral landmarks including frontozygomatic-suture point (FZS) and lateral-orbit point (LO), and the midline landmarks including Cg, ANS, upper dental midpoint (UDM), lower dental midpoint (LDM), and Me using V-Ceph 8.0 program. For test, Examiner-1 and Examiner-2 (3-year and 1-year orthodontic resident) and Cascaded-CNN model marked the landmarks. After point-to-point errors of landmark identification, successful detection rate (SDR, percentage within 1-, 2-, and 3-mm ranges), and distance and direction of the midline landmark deviation from the midsagittal line (ANS-mid, UDM-mid, LDM-mid, and Me-mid) were measured, statistical analysis was performed.
Results: The cascaded-CNN algorithm showed clinically acceptable level of point-to-point error (1.26 mm vs. 1.57 mm in Examiner-1 and 1.75 mm in Examiner-2). Its average SDR within 2 mm range was 83.2% with high accuracy at the right LO (96.9%), left LO (97.1%), and UDM (96.9%). Its absolute measurement errors were less than 1 mm in ANS-mid, UDM-mid, and LDM-mid compared to the gold standard.
Conclusion: The cascaded-CNN model might be considered an effective tool for auto-identification of the midline landmarks and quantification of the midline deviation in PA cephalograms of adult patients, regardless of variations in image acquisition method.
Keywords: Artificial intelligence, convolutional neural network, posteroanterior cephalograms
Korean J Orthod
First Published Date October 18, 2023
Copyright © The Korean Association of Orthodontists.
1st author: Sung-Hoon Han,a, 1st author: Jisup Lim,c, Jun-Sik Kim,d, Jin-Hyoung Cho,e, Mihee Hong,f, Minji Kim,g, Su-Jung Kim,h, Yoon-Ji Kim,i, Young Ho Kim,j, Sung-Hoon Lim,k, Sang Jin Sung,i, Kyung-Hwa Kang,a, Seung-Hak Baek,l, corresponding author: Sung-Kwon Choi,a, corresponding author: Namkug Kim,b
aDepartment of Orthodontics, School of Dentistry, Wonkwang University, Iksan-si, South Korea.
bDepartment of Convergence Medicine, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea.
cDepartment of Convergence Medicine, University of Ulsan, College of Medicine, Asan Medical Center, Seoul, Republic of Korea
dGraduate student (MSc), Department of Biomedical Engineering, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, Republic of Korea
eDepartment of Orthodontics, School of Dentistry, Chonnam National University, Gwangju, South Korea.
fDepartment of Orthodontics, School of Dentistry, Kyungpook National University, Daegu, South Korea.
gDepartment of Orthodontics, College of Medicine, Ewha Womans University, Seoul, South Korea.
hDepartment of Orthodontics, School of Dentistry, Kyung Hee University, Seoul, South Korea.
iDepartment of Orthodontics, Asan Medical Center, College of Medicine, University of Ulsan, Seoul, South Korea.
jDepartment of Orthodontics, Institute of Oral Health Science, Ajou University School of Medicine, Suwon-si, South Korea.
kDepartment of Orthodontics, College of Dentistry, Chosun University, Gwangju, South Korea.
lDepartment of Orthodontics, School of Dentistry, Dental Research Institute, Seoul National University, Seoul, South Korea.
Correspondence to:Sung-Kwon Choi, Department of Orthodontics, School of Dentistry, Wonkwang University, 460 Iksandae-ro, Iksan-si, Jeollabuk-do 54538, South Korea; Tel: +82-63-859-2962; e-mail, chsk6206@wku.ac.kr
or
Namkug Kim, Department of Convergence Medicine, Asan Medical Institute of Convergence Science and Technology, Asan Medical Center, University of Ulsan College of Medicine, 88 Olympic-Ro 43-Gil, Songpa-Gu, Seoul 05505, South Korea; Tel: +82-2-3010-6573; e-mail, namkugkim@gmail.com.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Objective: To quantify the effects of midline-related landmark identification in posteroanterior (PA) cephalogram images by cascaded convolutional neural network (CNN) algorithm on the midline deviation measurements.
Methods: A total of 2,903 PA cephalogram images obtained from nine university hospitals were divided into the training-set (n=2,150), internal validation-set (n=376), and test-set (n=377). As gold standard, two orthodontic professors marked the bilateral landmarks including frontozygomatic-suture point (FZS) and lateral-orbit point (LO), and the midline landmarks including Cg, ANS, upper dental midpoint (UDM), lower dental midpoint (LDM), and Me using V-Ceph 8.0 program. For test, Examiner-1 and Examiner-2 (3-year and 1-year orthodontic resident) and Cascaded-CNN model marked the landmarks. After point-to-point errors of landmark identification, successful detection rate (SDR, percentage within 1-, 2-, and 3-mm ranges), and distance and direction of the midline landmark deviation from the midsagittal line (ANS-mid, UDM-mid, LDM-mid, and Me-mid) were measured, statistical analysis was performed.
Results: The cascaded-CNN algorithm showed clinically acceptable level of point-to-point error (1.26 mm vs. 1.57 mm in Examiner-1 and 1.75 mm in Examiner-2). Its average SDR within 2 mm range was 83.2% with high accuracy at the right LO (96.9%), left LO (97.1%), and UDM (96.9%). Its absolute measurement errors were less than 1 mm in ANS-mid, UDM-mid, and LDM-mid compared to the gold standard.
Conclusion: The cascaded-CNN model might be considered an effective tool for auto-identification of the midline landmarks and quantification of the midline deviation in PA cephalograms of adult patients, regardless of variations in image acquisition method.
Keywords: Artificial intelligence, convolutional neural network, posteroanterior cephalograms
PDF
Standard view
Export citation
Share 
