An Early Diagnosis of Oral Cancer based on Three-Dimensional Convolutional Neural Networks

Shipu Xu; Chang Liu; Yongshuo Zong; Sirui Chen; Yiwen Lu; Longzhi Yang; Eddie Y. K. Ng; Yongtong Wang; Yunsheng Wang; Yong Liu; Wenwen Hu; Chenxi Zhang

doi:10.1109/ACCESS.2019.2950286

An Early Diagnosis of Oral Cancer based on Three-Dimensional Convolutional Neural Networks

Shipu Xu, Chang Liu, Yongshuo Zong, Sirui Chen, Yiwen Lu, Longzhi Yang, Eddie Y. K. Ng, Yongtong Wang, Yunsheng Wang, Yong Liu, Wenwen Hu, Chenxi Zhang

School of Computer Science

Research output: Contribution to journal › Article › peer-review

78 Citations (Scopus)

79 Downloads (Pure)

Abstract

Three-dimensional convolutional neural networks (3DCNNs), a rapidly evolving modality of deep learning, has gained popularity in many fields. For oral cancers, CT images are traditionally processed using two-dimensional input, without considering information between lesion slices. In this paper, we established a 3DCNNs-based image processing algorithm for the early diagnosis of oral cancers, which was compared with a 2DCNNs-based algorithm. The 3D and 2D CNNs were constructed using the same hierarchical structure to profile oral tumors as benign or malignant. Our results showed that 3DCNNs with dynamic characteristics of the enhancement rate image performed better than 2DCNNS with single enhancement sequence for the discrimination of oral cancer lesions. Our data indicate that spatial features and spatial dynamics extracted from 3DCNNs may inform future design of CT-assisted diagnosis system.

Original language	English
Pages (from-to)	158603-158611
Journal	IEEE Access
Volume	7
Early online date	30 Oct 2019
DOIs	https://doi.org/10.1109/ACCESS.2019.2950286
Publication status	Published - 12 Nov 2019

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

10.1109/ACCESS.2019.2950286Licence: CC BY

08887444Final published version, 1.55 MBLicence: CC BY

Cite this

@article{5868a488e4c144c680036b53295421fe,

title = "An Early Diagnosis of Oral Cancer based on Three-Dimensional Convolutional Neural Networks",

abstract = "Three-dimensional convolutional neural networks (3DCNNs), a rapidly evolving modality of deep learning, has gained popularity in many fields. For oral cancers, CT images are traditionally processed using two-dimensional input, without considering information between lesion slices. In this paper, we established a 3DCNNs-based image processing algorithm for the early diagnosis of oral cancers, which was compared with a 2DCNNs-based algorithm. The 3D and 2D CNNs were constructed using the same hierarchical structure to profile oral tumors as benign or malignant. Our results showed that 3DCNNs with dynamic characteristics of the enhancement rate image performed better than 2DCNNS with single enhancement sequence for the discrimination of oral cancer lesions. Our data indicate that spatial features and spatial dynamics extracted from 3DCNNs may inform future design of CT-assisted diagnosis system.",

author = "Shipu Xu and Chang Liu and Yongshuo Zong and Sirui Chen and Yiwen Lu and Longzhi Yang and Ng, \{Eddie Y. K.\} and Yongtong Wang and Yunsheng Wang and Yong Liu and Wenwen Hu and Chenxi Zhang",

year = "2019",

month = nov,

day = "12",

doi = "10.1109/ACCESS.2019.2950286",

language = "English",

volume = "7",

pages = "158603--158611",

journal = "IEEE Access",

issn = "2169-3536",

publisher = "IEEE",

}

TY - JOUR

T1 - An Early Diagnosis of Oral Cancer based on Three-Dimensional Convolutional Neural Networks

AU - Xu, Shipu

AU - Liu, Chang

AU - Zong, Yongshuo

AU - Chen, Sirui

AU - Lu, Yiwen

AU - Yang, Longzhi

AU - Ng, Eddie Y. K.

AU - Wang, Yongtong

AU - Wang, Yunsheng

AU - Liu, Yong

AU - Hu, Wenwen

AU - Zhang, Chenxi

PY - 2019/11/12

Y1 - 2019/11/12

N2 - Three-dimensional convolutional neural networks (3DCNNs), a rapidly evolving modality of deep learning, has gained popularity in many fields. For oral cancers, CT images are traditionally processed using two-dimensional input, without considering information between lesion slices. In this paper, we established a 3DCNNs-based image processing algorithm for the early diagnosis of oral cancers, which was compared with a 2DCNNs-based algorithm. The 3D and 2D CNNs were constructed using the same hierarchical structure to profile oral tumors as benign or malignant. Our results showed that 3DCNNs with dynamic characteristics of the enhancement rate image performed better than 2DCNNS with single enhancement sequence for the discrimination of oral cancer lesions. Our data indicate that spatial features and spatial dynamics extracted from 3DCNNs may inform future design of CT-assisted diagnosis system.

AB - Three-dimensional convolutional neural networks (3DCNNs), a rapidly evolving modality of deep learning, has gained popularity in many fields. For oral cancers, CT images are traditionally processed using two-dimensional input, without considering information between lesion slices. In this paper, we established a 3DCNNs-based image processing algorithm for the early diagnosis of oral cancers, which was compared with a 2DCNNs-based algorithm. The 3D and 2D CNNs were constructed using the same hierarchical structure to profile oral tumors as benign or malignant. Our results showed that 3DCNNs with dynamic characteristics of the enhancement rate image performed better than 2DCNNS with single enhancement sequence for the discrimination of oral cancer lesions. Our data indicate that spatial features and spatial dynamics extracted from 3DCNNs may inform future design of CT-assisted diagnosis system.

UR - https://www.scopus.com/pages/publications/85078701749

U2 - 10.1109/ACCESS.2019.2950286

DO - 10.1109/ACCESS.2019.2950286

M3 - Article

SN - 2169-3536

VL - 7

SP - 158603

EP - 158611

JO - IEEE Access

JF - IEEE Access

ER -

An Early Diagnosis of Oral Cancer based on Three-Dimensional Convolutional Neural Networks

Abstract

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this