Publication Details

Deep Learning for Cranioplasty in Clinical Practice: Going from Synthetic to Real Patient Data

KODYM, O.; ŠPANĚL, M.; HEROUT, A. Deep Learning for Cranioplasty in Clinical Practice: Going from Synthetic to Real Patient Data. COMPUTERS IN BIOLOGY AND MEDICINE, 2021, vol. 137, no. 104766, p. 1-10. ISSN: 0010-4825.
Czech title
Hluboké učení pro kranioplastiku v klinické praxi: od syntetických dat k reálným pacientům
Type
journal article
Language
English
Authors
Kodym Oldřich, Ing., Ph.D.
Španěl Michal, doc. Ing., Ph.D. (DCGM)
Herout Adam, prof. Ing., Ph.D. (DCGM)
URL
Keywords


Cranioplasty; Skull Reconstruction; Cranial Implant Design; 3D Convolutional
Neural
Networks

Abstract

Correct virtual reconstruction of a de-
fective skull is a prerequisite for successful cranioplasty
and its automatization has the potential for accelerat-
ing and standardizing the clinical workflow. This work
provides a deep learning-based method for the recon-
struction of a skull shape and cranial implant design
on clinical data of patients indicated for cranioplasty.
The method is based on a cascade of multi-branch vol-
umetric CNNs that enables simultaneous training on
two different types of cranioplasty ground-truth data:
the skull patch, which represents the exact shape of the
missing part of the original skull, and which can be eas-
ily created artificially from healthy skulls, and expert-
designed cranial implant shapes that are much harder
to acquire. The proposed method reaches an average
surface distance of the reconstructed skull patches of
0.67 mm on a clinical test set of 75 defective skulls. It
also achieves a 12% reduction of a newly proposed de-
fect border Gaussian curvature error metric, compared
to a baseline model trained on synthetic data only. Ad-
ditionally, it produces directly 3D printable cranial im-
plant shapes with a Dice coefficient 0.88 and a surface
error of 0.65 mm. The outputs of the proposed skull
reconstruction method reach good quality and can be
considered for use in semi- or fully automatic clinical
cranial implant design workflows.

Published
2021
Pages
1–10
Journal
COMPUTERS IN BIOLOGY AND MEDICINE, vol. 137, no. 104766, ISSN 0010-4825
DOI
UT WoS
000704338500006
EID Scopus
BibTeX
@article{BUT175781,
  author="Oldřich {Kodym} and Michal {Španěl} and Adam {Herout}",
  title="Deep Learning for Cranioplasty in Clinical Practice: Going from Synthetic to Real Patient Data",
  journal="COMPUTERS IN BIOLOGY AND MEDICINE",
  year="2021",
  volume="137",
  number="104766",
  pages="1--10",
  doi="10.1016/j.compbiomed.2021.104766",
  issn="0010-4825",
  url="https://www.sciencedirect.com/science/article/abs/pii/S0010482521005606?via%3Dihub"
}
Files
Back to top