Automatic craniomaxillofacial landmark digitization via segmentation-guided partially-joint regression forest model

Jun Zhang, Yaozong Gao, Li Wang, Zhen Tang, James J. Xia, Dinggang Shen

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

Craniomaxillofacial (CMF) deformities involve congenital and acquired deformities of the head and face. Landmark digitization is a critical step in quantifying CMF deformities. In current clinical practice, CMF landmarks have to be manually digitized on 3D models, which is time-consuming. To date, there is no clinically acceptable method that allows automatic landmark digitization, due to morphological variations among different patients and artifacts of cone-beam computed tomography (CBCT) images. To address these challenges, we propose a segmentation-guided partially-joint regression forest model that can automatically digitizes CMF landmarks. In this model, a regression voting strategy is first adopted to localize landmarks by aggregating evidences from context locations, thus potentially relieving the problem caused by image artifacts near the landmark. Second, segmentation is also utilized to resolve inconsistent landmark appearances that are caused by morphological variations among different patients, especially on the teeth. Third, a partially-joint model is proposed to separately localize landmarks based on coherence of landmark positions to improve digitization reliability. The experimental results show that the accuracy of automatically digitized landmarks using our approach is clinically acceptable.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PublisherSpringer Verlag
Pages661-668
Number of pages8
Volume9351
ISBN (Print)9783319245737
DOIs
Publication statusPublished - 2015
Externally publishedYes
Event18th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2015 - Munich, Germany
Duration: 2015 Oct 52015 Oct 9

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume9351
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other18th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2015
CountryGermany
CityMunich
Period15/10/515/10/9

Fingerprint

Digitization
Analog to digital conversion
Landmarks
Segmentation
Regression
Model
Tomography
Cones
Joint Model
Computed Tomography
Voting
3D Model
Inconsistent
Resolve
Cone
Face

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Zhang, J., Gao, Y., Wang, L., Tang, Z., Xia, J. J., & Shen, D. (2015). Automatic craniomaxillofacial landmark digitization via segmentation-guided partially-joint regression forest model. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9351, pp. 661-668). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9351). Springer Verlag. https://doi.org/10.1007/978-3-319-24574-4_79

Automatic craniomaxillofacial landmark digitization via segmentation-guided partially-joint regression forest model. / Zhang, Jun; Gao, Yaozong; Wang, Li; Tang, Zhen; Xia, James J.; Shen, Dinggang.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9351 Springer Verlag, 2015. p. 661-668 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9351).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Zhang, J, Gao, Y, Wang, L, Tang, Z, Xia, JJ & Shen, D 2015, Automatic craniomaxillofacial landmark digitization via segmentation-guided partially-joint regression forest model. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 9351, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9351, Springer Verlag, pp. 661-668, 18th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2015, Munich, Germany, 15/10/5. https://doi.org/10.1007/978-3-319-24574-4_79
Zhang J, Gao Y, Wang L, Tang Z, Xia JJ, Shen D. Automatic craniomaxillofacial landmark digitization via segmentation-guided partially-joint regression forest model. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9351. Springer Verlag. 2015. p. 661-668. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-24574-4_79
Zhang, Jun ; Gao, Yaozong ; Wang, Li ; Tang, Zhen ; Xia, James J. ; Shen, Dinggang. / Automatic craniomaxillofacial landmark digitization via segmentation-guided partially-joint regression forest model. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 9351 Springer Verlag, 2015. pp. 661-668 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{f5c1a78c3bdc4682bb6136964880834e,
title = "Automatic craniomaxillofacial landmark digitization via segmentation-guided partially-joint regression forest model",
abstract = "Craniomaxillofacial (CMF) deformities involve congenital and acquired deformities of the head and face. Landmark digitization is a critical step in quantifying CMF deformities. In current clinical practice, CMF landmarks have to be manually digitized on 3D models, which is time-consuming. To date, there is no clinically acceptable method that allows automatic landmark digitization, due to morphological variations among different patients and artifacts of cone-beam computed tomography (CBCT) images. To address these challenges, we propose a segmentation-guided partially-joint regression forest model that can automatically digitizes CMF landmarks. In this model, a regression voting strategy is first adopted to localize landmarks by aggregating evidences from context locations, thus potentially relieving the problem caused by image artifacts near the landmark. Second, segmentation is also utilized to resolve inconsistent landmark appearances that are caused by morphological variations among different patients, especially on the teeth. Third, a partially-joint model is proposed to separately localize landmarks based on coherence of landmark positions to improve digitization reliability. The experimental results show that the accuracy of automatically digitized landmarks using our approach is clinically acceptable.",
author = "Jun Zhang and Yaozong Gao and Li Wang and Zhen Tang and Xia, {James J.} and Dinggang Shen",
year = "2015",
doi = "10.1007/978-3-319-24574-4_79",
language = "English",
isbn = "9783319245737",
volume = "9351",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "661--668",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Automatic craniomaxillofacial landmark digitization via segmentation-guided partially-joint regression forest model

AU - Zhang, Jun

AU - Gao, Yaozong

AU - Wang, Li

AU - Tang, Zhen

AU - Xia, James J.

AU - Shen, Dinggang

PY - 2015

Y1 - 2015

N2 - Craniomaxillofacial (CMF) deformities involve congenital and acquired deformities of the head and face. Landmark digitization is a critical step in quantifying CMF deformities. In current clinical practice, CMF landmarks have to be manually digitized on 3D models, which is time-consuming. To date, there is no clinically acceptable method that allows automatic landmark digitization, due to morphological variations among different patients and artifacts of cone-beam computed tomography (CBCT) images. To address these challenges, we propose a segmentation-guided partially-joint regression forest model that can automatically digitizes CMF landmarks. In this model, a regression voting strategy is first adopted to localize landmarks by aggregating evidences from context locations, thus potentially relieving the problem caused by image artifacts near the landmark. Second, segmentation is also utilized to resolve inconsistent landmark appearances that are caused by morphological variations among different patients, especially on the teeth. Third, a partially-joint model is proposed to separately localize landmarks based on coherence of landmark positions to improve digitization reliability. The experimental results show that the accuracy of automatically digitized landmarks using our approach is clinically acceptable.

AB - Craniomaxillofacial (CMF) deformities involve congenital and acquired deformities of the head and face. Landmark digitization is a critical step in quantifying CMF deformities. In current clinical practice, CMF landmarks have to be manually digitized on 3D models, which is time-consuming. To date, there is no clinically acceptable method that allows automatic landmark digitization, due to morphological variations among different patients and artifacts of cone-beam computed tomography (CBCT) images. To address these challenges, we propose a segmentation-guided partially-joint regression forest model that can automatically digitizes CMF landmarks. In this model, a regression voting strategy is first adopted to localize landmarks by aggregating evidences from context locations, thus potentially relieving the problem caused by image artifacts near the landmark. Second, segmentation is also utilized to resolve inconsistent landmark appearances that are caused by morphological variations among different patients, especially on the teeth. Third, a partially-joint model is proposed to separately localize landmarks based on coherence of landmark positions to improve digitization reliability. The experimental results show that the accuracy of automatically digitized landmarks using our approach is clinically acceptable.

UR - http://www.scopus.com/inward/record.url?scp=84951744295&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84951744295&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-24574-4_79

DO - 10.1007/978-3-319-24574-4_79

M3 - Conference contribution

AN - SCOPUS:84951744295

SN - 9783319245737

VL - 9351

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 661

EP - 668

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

PB - Springer Verlag

ER -