Automatic three-dimensional cephalometric annotation system using three-dimensional convolutional neural networks: a developmental trial

Sung Ho Kang, Kiwan Jeon, Hak Jin Kim, Jin Keun Seo, Sang Hwy Lee

Research output: Contribution to journalArticle

Abstract

Automatic annotation for three-dimensional (3D) cephalometric analysis has been limited by computational complexity and computing performance. The purpose of this study was to evaluate the accuracy of our newly-developed automatic 3D cephalometric annotation system using a deep learning algorithm. Our model system mainly consisted of a 3D convolutional neural network and image data resampling. Discrepancies between the referenced and predicted coordinate values in three axes and in 3D distance were calculated to yield prediction errors of 3.26, 3.18, and 4.81 mm (for three axes) and 7.61 mm (for 3D). Moreover, there was no difference (p > 0.05) among the landmarks of three groups (midsagittal plane, horizontal plane and mandible). Although our 3D convolutional neural network-based annotation system could not achieve the level of accuracy demanded by immediate clinical applications, it can nevertheless serve as an initial approximate guide to landmarks, thus reducing the time needed for annotation.

Fingerprint

Cephalometry
Neural networks
Mandible
Learning algorithms
Computational complexity
Learning

All Science Journal Classification (ASJC) codes

  • Computational Mechanics
  • Biomedical Engineering
  • Radiology Nuclear Medicine and imaging
  • Computer Science Applications

Cite this

@article{efeb548aaf18492dbe0671408c432b59,
title = "Automatic three-dimensional cephalometric annotation system using three-dimensional convolutional neural networks: a developmental trial",
abstract = "Automatic annotation for three-dimensional (3D) cephalometric analysis has been limited by computational complexity and computing performance. The purpose of this study was to evaluate the accuracy of our newly-developed automatic 3D cephalometric annotation system using a deep learning algorithm. Our model system mainly consisted of a 3D convolutional neural network and image data resampling. Discrepancies between the referenced and predicted coordinate values in three axes and in 3D distance were calculated to yield prediction errors of 3.26, 3.18, and 4.81 mm (for three axes) and 7.61 mm (for 3D). Moreover, there was no difference (p > 0.05) among the landmarks of three groups (midsagittal plane, horizontal plane and mandible). Although our 3D convolutional neural network-based annotation system could not achieve the level of accuracy demanded by immediate clinical applications, it can nevertheless serve as an initial approximate guide to landmarks, thus reducing the time needed for annotation.",
author = "Kang, {Sung Ho} and Kiwan Jeon and Kim, {Hak Jin} and Seo, {Jin Keun} and Lee, {Sang Hwy}",
year = "2019",
month = "1",
day = "1",
doi = "10.1080/21681163.2019.1674696",
language = "English",
journal = "Computer Methods in Biomechanics and Biomedical Engineering: Imaging and Visualization",
issn = "2168-1163",
publisher = "Taylor and Francis Ltd.",

}

TY - JOUR

T1 - Automatic three-dimensional cephalometric annotation system using three-dimensional convolutional neural networks

T2 - a developmental trial

AU - Kang, Sung Ho

AU - Jeon, Kiwan

AU - Kim, Hak Jin

AU - Seo, Jin Keun

AU - Lee, Sang Hwy

PY - 2019/1/1

Y1 - 2019/1/1

N2 - Automatic annotation for three-dimensional (3D) cephalometric analysis has been limited by computational complexity and computing performance. The purpose of this study was to evaluate the accuracy of our newly-developed automatic 3D cephalometric annotation system using a deep learning algorithm. Our model system mainly consisted of a 3D convolutional neural network and image data resampling. Discrepancies between the referenced and predicted coordinate values in three axes and in 3D distance were calculated to yield prediction errors of 3.26, 3.18, and 4.81 mm (for three axes) and 7.61 mm (for 3D). Moreover, there was no difference (p > 0.05) among the landmarks of three groups (midsagittal plane, horizontal plane and mandible). Although our 3D convolutional neural network-based annotation system could not achieve the level of accuracy demanded by immediate clinical applications, it can nevertheless serve as an initial approximate guide to landmarks, thus reducing the time needed for annotation.

AB - Automatic annotation for three-dimensional (3D) cephalometric analysis has been limited by computational complexity and computing performance. The purpose of this study was to evaluate the accuracy of our newly-developed automatic 3D cephalometric annotation system using a deep learning algorithm. Our model system mainly consisted of a 3D convolutional neural network and image data resampling. Discrepancies between the referenced and predicted coordinate values in three axes and in 3D distance were calculated to yield prediction errors of 3.26, 3.18, and 4.81 mm (for three axes) and 7.61 mm (for 3D). Moreover, there was no difference (p > 0.05) among the landmarks of three groups (midsagittal plane, horizontal plane and mandible). Although our 3D convolutional neural network-based annotation system could not achieve the level of accuracy demanded by immediate clinical applications, it can nevertheless serve as an initial approximate guide to landmarks, thus reducing the time needed for annotation.

UR - http://www.scopus.com/inward/record.url?scp=85074321214&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85074321214&partnerID=8YFLogxK

U2 - 10.1080/21681163.2019.1674696

DO - 10.1080/21681163.2019.1674696

M3 - Article

AN - SCOPUS:85074321214

JO - Computer Methods in Biomechanics and Biomedical Engineering: Imaging and Visualization

JF - Computer Methods in Biomechanics and Biomedical Engineering: Imaging and Visualization

SN - 2168-1163

ER -