High-Resolution Cranial Implant Prediction via Patch-Wise Training

Yuan Jin; Jianning Li; Jan Egger

doi:10.1007/978-3-030-64327-0_11

High-Resolution Cranial Implant Prediction via Patch-Wise Training

Yuan Jin, Jianning Li, Jan Egger

Institute of Computer Graphics and Vision (7100)

Research output: Chapter in Book/Report/Conference proceeding › Conference paper › peer-review

Abstract

In this study, we proposed two methods for AutoImplant (https://autoimplant.grand-challenge.org/) - the cranial implant design challenge. The shape of the implant is predicted based on the inputted defective skull. This task can be accomplished either by directly predicting the implant with the defective skull, or indirectly rebuilding the complete skull and then taking the difference between the defective and complete skulls. In our work, a deep learning model is applied to automatically predict the implant. In order to solve the problem that high resolution images can often not be directly inputted to the deep learning model, two proposed methods of resize and patch-based are examined. On the test set, the proposed resize method achieves an average dice similarity score (DSC) of 0.7350 and a Hausdorff distance (HD) of 7.2425 mm, while the proposed patch-based method achieves an average DSC of 0.8887 and a HD of 5.5339 mm.

Original language	English
Title of host publication	Towards the Automatization of Cranial Implant Design in Cranioplasty - First Challenge, AutoImplant 2020, Held in Conjunction with MICCAI 2020, Proceedings
Editors	Jianning Li, Jan Egger
Publisher	Springer, Cham
Pages	94 -103
Number of pages	10
Volume	12439
Edition	1
ISBN (Print)	9783030643263
DOIs	https://doi.org/10.1007/978-3-030-64327-0_11
Publication status	E-pub ahead of print - 4 Dec 2020
Event	1st Automatization of Cranial Implant Design in Cranioplasty Challenge: AutoImplant 2020 - Virtual, Lima, Peru Duration: 8 Oct 2020 → … https://autoimplant.grand-challenge.org/

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	12439 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	1st Automatization of Cranial Implant Design in Cranioplasty Challenge
Country/Territory	Peru
City	Virtual, Lima
Period	8/10/20 → …
Internet address	https://autoimplant.grand-challenge.org/

Keywords

AutoImplant
Cranioplasty
Deep learning
Shape completion
Super-resolution

ASJC Scopus subject areas

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-030-64327-0_11

Cite this

Jin, Y., Li, J., & Egger, J. (2020). High-Resolution Cranial Implant Prediction via Patch-Wise Training. In J. Li, & J. Egger (Eds.), Towards the Automatization of Cranial Implant Design in Cranioplasty - First Challenge, AutoImplant 2020, Held in Conjunction with MICCAI 2020, Proceedings (1 ed., Vol. 12439, pp. 94 -103). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12439 LNCS). Springer, Cham. Advance online publication. https://doi.org/10.1007/978-3-030-64327-0_11

High-Resolution Cranial Implant Prediction via Patch-Wise Training. / Jin, Yuan; Li, Jianning; Egger, Jan.
Towards the Automatization of Cranial Implant Design in Cranioplasty - First Challenge, AutoImplant 2020, Held in Conjunction with MICCAI 2020, Proceedings. ed. / Jianning Li; Jan Egger. Vol. 12439 1. ed. Springer, Cham, 2020. p. 94 -103 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 12439 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference paper › peer-review

Jin, Y, Li, J & Egger, J 2020, High-Resolution Cranial Implant Prediction via Patch-Wise Training. in J Li & J Egger (eds), Towards the Automatization of Cranial Implant Design in Cranioplasty - First Challenge, AutoImplant 2020, Held in Conjunction with MICCAI 2020, Proceedings. 1 edn, vol. 12439, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 12439 LNCS, Springer, Cham, pp. 94 -103, 1st Automatization of Cranial Implant Design in Cranioplasty Challenge, Virtual, Lima, Peru, 8/10/20. https://doi.org/10.1007/978-3-030-64327-0_11

Jin Y, Li J, Egger J. High-Resolution Cranial Implant Prediction via Patch-Wise Training. In Li J, Egger J, editors, Towards the Automatization of Cranial Implant Design in Cranioplasty - First Challenge, AutoImplant 2020, Held in Conjunction with MICCAI 2020, Proceedings. 1 ed. Vol. 12439. Springer, Cham. 2020. p. 94 -103. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). Epub 2020 Dec 4. doi: 10.1007/978-3-030-64327-0_11

Jin, Yuan ; Li, Jianning ; Egger, Jan. / High-Resolution Cranial Implant Prediction via Patch-Wise Training. Towards the Automatization of Cranial Implant Design in Cranioplasty - First Challenge, AutoImplant 2020, Held in Conjunction with MICCAI 2020, Proceedings. editor / Jianning Li ; Jan Egger. Vol. 12439 1. ed. Springer, Cham, 2020. pp. 94 -103 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{68e718243b8b4c79b13e065478492a9c,

title = "High-Resolution Cranial Implant Prediction via Patch-Wise Training",

abstract = "In this study, we proposed two methods for AutoImplant (https://autoimplant.grand-challenge.org/) - the cranial implant design challenge. The shape of the implant is predicted based on the inputted defective skull. This task can be accomplished either by directly predicting the implant with the defective skull, or indirectly rebuilding the complete skull and then taking the difference between the defective and complete skulls. In our work, a deep learning model is applied to automatically predict the implant. In order to solve the problem that high resolution images can often not be directly inputted to the deep learning model, two proposed methods of resize and patch-based are examined. On the test set, the proposed resize method achieves an average dice similarity score (DSC) of 0.7350 and a Hausdorff distance (HD) of 7.2425 mm, while the proposed patch-based method achieves an average DSC of 0.8887 and a HD of 5.5339 mm.",

keywords = "AutoImplant, Cranioplasty, Deep learning, Shape completion, Super-resolution",

author = "Yuan Jin and Jianning Li and Jan Egger",

year = "2020",

month = dec,

day = "4",

doi = "10.1007/978-3-030-64327-0_11",

language = "English",

isbn = "9783030643263",

volume = "12439",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer, Cham",

pages = "94 --103",

editor = "Jianning Li and Jan Egger",

booktitle = "Towards the Automatization of Cranial Implant Design in Cranioplasty - First Challenge, AutoImplant 2020, Held in Conjunction with MICCAI 2020, Proceedings",

edition = "1",

note = "1st Automatization of Cranial Implant Design in Cranioplasty Challenge : AutoImplant 2020 ; Conference date: 08-10-2020",

url = "https://autoimplant.grand-challenge.org/",

}

TY - GEN

T1 - High-Resolution Cranial Implant Prediction via Patch-Wise Training

AU - Jin, Yuan

AU - Li, Jianning

AU - Egger, Jan

PY - 2020/12/4

Y1 - 2020/12/4

N2 - In this study, we proposed two methods for AutoImplant (https://autoimplant.grand-challenge.org/) - the cranial implant design challenge. The shape of the implant is predicted based on the inputted defective skull. This task can be accomplished either by directly predicting the implant with the defective skull, or indirectly rebuilding the complete skull and then taking the difference between the defective and complete skulls. In our work, a deep learning model is applied to automatically predict the implant. In order to solve the problem that high resolution images can often not be directly inputted to the deep learning model, two proposed methods of resize and patch-based are examined. On the test set, the proposed resize method achieves an average dice similarity score (DSC) of 0.7350 and a Hausdorff distance (HD) of 7.2425 mm, while the proposed patch-based method achieves an average DSC of 0.8887 and a HD of 5.5339 mm.

AB - In this study, we proposed two methods for AutoImplant (https://autoimplant.grand-challenge.org/) - the cranial implant design challenge. The shape of the implant is predicted based on the inputted defective skull. This task can be accomplished either by directly predicting the implant with the defective skull, or indirectly rebuilding the complete skull and then taking the difference between the defective and complete skulls. In our work, a deep learning model is applied to automatically predict the implant. In order to solve the problem that high resolution images can often not be directly inputted to the deep learning model, two proposed methods of resize and patch-based are examined. On the test set, the proposed resize method achieves an average dice similarity score (DSC) of 0.7350 and a Hausdorff distance (HD) of 7.2425 mm, while the proposed patch-based method achieves an average DSC of 0.8887 and a HD of 5.5339 mm.

KW - AutoImplant

KW - Cranioplasty

KW - Deep learning

KW - Shape completion

KW - Super-resolution

UR - http://www.scopus.com/inward/record.url?scp=85097829680&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-64327-0_11

DO - 10.1007/978-3-030-64327-0_11

M3 - Conference paper

SN - 9783030643263

VL - 12439

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

SP - 94

EP - 103

BT - Towards the Automatization of Cranial Implant Design in Cranioplasty - First Challenge, AutoImplant 2020, Held in Conjunction with MICCAI 2020, Proceedings

A2 - Li, Jianning

A2 - Egger, Jan

PB - Springer, Cham

T2 - 1st Automatization of Cranial Implant Design in Cranioplasty Challenge

Y2 - 8 October 2020

ER -

High-Resolution Cranial Implant Prediction via Patch-Wise Training

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this