EfficientNet-Based Model With Test Time Augmentation for Cancer Detection

Jiahao Zhang; Yang Jiang; Rao Huang; Jiacheng Shi

doi:10.1109/ICBAIE52039.2021.9389825

EfficientNet-Based Model With Test Time Augmentation for Cancer Detection

Jiahao Zhang, Yang Jiang, Rao Huang, Jiacheng Shi

City University of Hong Kong

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

13 Citations (Scopus)

Abstract

Skin cancer is the most common type of cancer and among various kinds of skin cancer, melanoma causes the most deaths. In clinical practice, contextual information from every one of a patient's moles help dermatologists make better judgments about whether a particular one is a lesion. In this paper, we proposed an EfficientNet based deep learning method to identify melanoma in skin lesion images. Our method takes into consideration all skin lesion images from a patient and employs effective data augmentation during training and test time augmentation during inference to improve classification accuracy. On the SIIM-ISIC Melanoma Classification dataset, our method achieved 0.901 Auc-Roc scores, outperforming other deep learning models such as VGG16 or Resnet50.

Original language	English
Title of host publication	2021 IEEE 2nd International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE)
Publisher	IEEE
Pages	548-551
ISBN (Electronic)	978-1-6654-1540-8
ISBN (Print)	978-0-7381-3122-1, 978-1-6654-4706-5
DOIs	https://doi.org/10.1109/ICBAIE52039.2021.9389825
Publication status	Published - Mar 2021
Event	2021 IEEE 2nd International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE 2021) - Nanchang, China Duration: 26 Mar 2021 → 28 Mar 2021

Publication series

Name	IEEE International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering, ICBAIE

Conference

Conference	2021 IEEE 2nd International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE 2021)
Place	China
City	Nanchang
Period	26/03/21 → 28/03/21

UN SDGs

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

SDG 3 Good Health and Well-being

Research Keywords

cancer detection
convolution neural network
image classification
ResNet network

Access Output

10.1109/ICBAIE52039.2021.9389825

Other Access Options

Check CityUHK Library

Permanent Link

Right click to copy link

Cite this

Zhang, J., Jiang, Y., Huang, R., & Shi, J. (2021). EfficientNet-Based Model With Test Time Augmentation for Cancer Detection. In 2021 IEEE 2nd International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE) (pp. 548-551). Article 9389825 (IEEE International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering, ICBAIE). IEEE. https://doi.org/10.1109/ICBAIE52039.2021.9389825

Zhang, Jiahao ; Jiang, Yang ; Huang, Rao et al. / EfficientNet-Based Model With Test Time Augmentation for Cancer Detection. 2021 IEEE 2nd International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE). IEEE, 2021. pp. 548-551 (IEEE International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering, ICBAIE).

@inproceedings{c285baa11b6148579e2dc1d16432faa3,

title = "EfficientNet-Based Model With Test Time Augmentation for Cancer Detection",

abstract = "Skin cancer is the most common type of cancer and among various kinds of skin cancer, melanoma causes the most deaths. In clinical practice, contextual information from every one of a patient's moles help dermatologists make better judgments about whether a particular one is a lesion. In this paper, we proposed an EfficientNet based deep learning method to identify melanoma in skin lesion images. Our method takes into consideration all skin lesion images from a patient and employs effective data augmentation during training and test time augmentation during inference to improve classification accuracy. On the SIIM-ISIC Melanoma Classification dataset, our method achieved 0.901 Auc-Roc scores, outperforming other deep learning models such as VGG16 or Resnet50.",

keywords = "cancer detection, convolution neural network, image classification, ResNet network",

author = "Jiahao Zhang and Yang Jiang and Rao Huang and Jiacheng Shi",

year = "2021",

month = mar,

doi = "10.1109/ICBAIE52039.2021.9389825",

language = "English",

isbn = "978-0-7381-3122-1",

series = "IEEE International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering, ICBAIE",

publisher = "IEEE",

pages = "548--551",

booktitle = "2021 IEEE 2nd International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE)",

address = "United States",

note = "2021 IEEE 2nd International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE 2021) ; Conference date: 26-03-2021 Through 28-03-2021",

}

Zhang, J, Jiang, Y, Huang, R & Shi, J 2021, EfficientNet-Based Model With Test Time Augmentation for Cancer Detection. in 2021 IEEE 2nd International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE)., 9389825, IEEE International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering, ICBAIE, IEEE, pp. 548-551, 2021 IEEE 2nd International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE 2021), Nanchang, China, 26/03/21. https://doi.org/10.1109/ICBAIE52039.2021.9389825

EfficientNet-Based Model With Test Time Augmentation for Cancer Detection. / Zhang, Jiahao; Jiang, Yang; Huang, Rao et al.
2021 IEEE 2nd International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE). IEEE, 2021. p. 548-551 9389825 (IEEE International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering, ICBAIE).

Research output: Chapters, Conference Papers, Creative and Literary Works › RGC 32 - Refereed conference paper (with host publication) › peer-review

TY - GEN

T1 - EfficientNet-Based Model With Test Time Augmentation for Cancer Detection

AU - Zhang, Jiahao

AU - Jiang, Yang

AU - Huang, Rao

AU - Shi, Jiacheng

PY - 2021/3

Y1 - 2021/3

N2 - Skin cancer is the most common type of cancer and among various kinds of skin cancer, melanoma causes the most deaths. In clinical practice, contextual information from every one of a patient's moles help dermatologists make better judgments about whether a particular one is a lesion. In this paper, we proposed an EfficientNet based deep learning method to identify melanoma in skin lesion images. Our method takes into consideration all skin lesion images from a patient and employs effective data augmentation during training and test time augmentation during inference to improve classification accuracy. On the SIIM-ISIC Melanoma Classification dataset, our method achieved 0.901 Auc-Roc scores, outperforming other deep learning models such as VGG16 or Resnet50.

AB - Skin cancer is the most common type of cancer and among various kinds of skin cancer, melanoma causes the most deaths. In clinical practice, contextual information from every one of a patient's moles help dermatologists make better judgments about whether a particular one is a lesion. In this paper, we proposed an EfficientNet based deep learning method to identify melanoma in skin lesion images. Our method takes into consideration all skin lesion images from a patient and employs effective data augmentation during training and test time augmentation during inference to improve classification accuracy. On the SIIM-ISIC Melanoma Classification dataset, our method achieved 0.901 Auc-Roc scores, outperforming other deep learning models such as VGG16 or Resnet50.

KW - cancer detection

KW - convolution neural network

KW - image classification

KW - ResNet network

UR - http://www.scopus.com/inward/record.url?scp=85104411429&partnerID=8YFLogxK

UR - https://www.scopus.com/record/pubmetrics.uri?eid=2-s2.0-85104411429&origin=recordpage

U2 - 10.1109/ICBAIE52039.2021.9389825

DO - 10.1109/ICBAIE52039.2021.9389825

M3 - RGC 32 - Refereed conference paper (with host publication)

SN - 978-0-7381-3122-1

SN - 978-1-6654-4706-5

T3 - IEEE International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering, ICBAIE

SP - 548

EP - 551

BT - 2021 IEEE 2nd International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE)

PB - IEEE

T2 - 2021 IEEE 2nd International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE 2021)

Y2 - 26 March 2021 through 28 March 2021

ER -

Zhang J, Jiang Y, Huang R, Shi J. EfficientNet-Based Model With Test Time Augmentation for Cancer Detection. In 2021 IEEE 2nd International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering (ICBAIE). IEEE. 2021. p. 548-551. 9389825. (IEEE International Conference on Big Data, Artificial Intelligence and Internet of Things Engineering, ICBAIE). doi: 10.1109/ICBAIE52039.2021.9389825

EfficientNet-Based Model With Test Time Augmentation for Cancer Detection

Abstract

Publication series

Conference

UN SDGs

Research Keywords

Access Output

Other Access Options

Permanent Link

Other Files and Links

Fingerprint

Cite this