Citation: Seol, Y.J.; Kim, Y.J.; Kim,
Y.S.; Cheon, Y.W.; Kim, K.G. A Study
on 3D Deep Learning-Based
Automatic Diagnosis of Nasal
Fractures. Sensors 2022, 22, 506.
https://doi.org/
10.3390/s22020506
Academic Editors: Kelvin K.L. Wong,
Dhanjoo N. Ghista, Andrew W.H. Ip
and Wenjun (Chris) Zhang
Received: 23 November 2021
Accepted: 4 January 2022
Published: 10 January 2022
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Communication
A Study on 3D Deep Learning-Based Automatic Diagnosis of
Nasal Fractures
Yu Jin Seol
1
, Young Jae Kim
2
, Yoon Sang Kim
3
, Young Woo Cheon
3,
* and Kwang Gi Kim
2,4,
*
1
Department of Biomedical Engineering, Gachon University, 191, Hambangmoe-ro, Yeonsu-gu,
Incheon 21936, Korea; tjfwlgns0518@gmail.com
2
Department of Biomedical Engineering, Gachon University College of Medicine, 38-13 Docjeom-ro 3 beon-gil,
Namdong-gu, Incheon 21565, Korea; youngjae@gachon.ac.kr
3
Department of Plastic and Reconstructive Surgery, Gachon University Gil Medical Center, College of
Medicine, Incheon 21565, Korea; yunsang0115@gmail.com
4
Department of Health Sciences and Technology, Gachon Advanced Institute for Health Sciences and
Technology (GAIHST), Gachon University, Seongnam-si 13120, Korea
* Correspondence: youngwooc@gmail.com (Y.W.C.); kimkg@gachon.ac.kr (K.G.K.)
Abstract:
This paper reported a study on the 3-dimensional deep-learning-based automatic diagnosis
of nasal fractures. (1) Background: The nasal bone is the most protuberant feature of the face;
therefore, it is highly vulnerable to facial trauma and its fractures are known as the most common
facial fractures worldwide. In addition, its adhesion causes rapid deformation, so a clear diagnosis
is needed early after fracture onset. (2) Methods: The collected computed tomography images
were reconstructed to isotropic voxel data including the whole region of the nasal bone, which are
represented in a fixed cubic volume. The configured 3-dimensional input data were then automatically
classified by the deep learning of residual neural networks (3D-ResNet34 and ResNet50) with the
spatial context information using a single network, whose performance was evaluated by 5-fold
cross-validation. (3) Results: The classification of nasal fractures with simple 3D-ResNet34 and
ResNet50 networks achieved areas under the receiver operating characteristic curve of 94.5% and
93.4% for binary classification, respectively, both indicating unprecedented high performance in the
task.
(4) Conclusions:
In this paper, it is presented the possibility of automatic nasal bone fracture
diagnosis using a 3-dimensional Resnet-based single classification network and it will improve the
diagnostic environment with future research.
Keywords:
artificial intelligence; computed aided diagnosis (CAD); 3D-classification; nasal fractures
1. Introduction
The nasal bone is the most prominent part of the facial skeleton, making it more vulner-
able to traumatic fractures. Therefore, nasal fractures are the most frequent facial fractures
worldwide and can be caused by relatively weak forces [
1
,
2
]. Furthermore, nasal fractures
due to facial trauma retain their state in a relatively short time, within 1–2 weeks, causing
nasal deformity with a high incidence of approximately 14 to 50 percent [
3
]. Rhinoplasty for
nasal deformity from a facial trauma is one of the most challenging problems for surgeons;
thus, nasal fractures require accurate diagnosis immediately after the onset and before defor-
mation occurs [
4
]. To perform sophisticated diagnosis of nasal fractures and determination
of the range and pattern of a fractured nose, radiologists have recently used computed
tomography (CT) images, mostly achieving excellent sensitivity and specificity [5].
The diagnosis of traumatic nasal fractures from facial CT analysis is an effective
tool for their early reconstruction. However, the process of CT image reading not only
burdens readers (radiologists) with the need to examine numerous CT images in great
detail but also makes it difficult to detect fine nasal bone defects in fractures, with high
subjectivity in the diagnostic results reported by each reader [
6
]. Thus, the diagnosis of
Sensors 2022, 22, 506. https://doi.org/10.3390/s22020506 https://www.mdpi.com/journal/sensors
