DSSAE: Deep Stacked Sparse Autoencoder Analytical Model for COVID-19 Diagnosis by Fractional Fourier Entropy

Shui Hua Wang; Xin Zhang; Yu Dong Zhang

doi:10.1145/3451357

DSSAE: Deep Stacked Sparse Autoencoder Analytical Model for COVID-19 Diagnosis by Fractional Fourier Entropy

Shui Hua Wang, Xin Zhang^*, Yu Dong Zhang^*

^*Corresponding author for this work

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

29 Citations (Scopus)

Abstract

(Aim) COVID-19 has caused more than 2.28 million deaths till 4/Feb/2021 while it is still spreading across the world. This study proposed a novel artificial intelligence model to diagnose COVID-19 based on chest CT images. (Methods) First, the two-dimensional fractional Fourier entropy was used to extract features. Second, a custom deep stacked sparse autoencoder (DSSAE) model was created to serve as the classifier. Third, an improved multiple-way data augmentation was proposed to resist overfitting. (Results) Our DSSAE model obtains a micro-averaged F1 score of 92.32% in handling a four-class problem (COVID-19, community-acquired pneumonia, secondary pulmonary tuberculosis, and healthy control). (Conclusion) Our method outperforms 10 state-of-the-art approaches.

Original language	English
Article number	2
Journal	ACM Transactions on Management Information Systems
Volume	13
Issue number	1
DOIs	https://doi.org/10.1145/3451357
Publication status	Published - 5 Oct 2021
Externally published	Yes

Keywords

COVID-19
Deep learning
autoencoder
fractional Fourier entropy

UN SDGs

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

Access to Document

10.1145/3451357

Cite this

@article{d0fc3a59d70f4b3db8e386558e62ef0b,

title = "DSSAE: Deep Stacked Sparse Autoencoder Analytical Model for COVID-19 Diagnosis by Fractional Fourier Entropy",

abstract = "(Aim) COVID-19 has caused more than 2.28 million deaths till 4/Feb/2021 while it is still spreading across the world. This study proposed a novel artificial intelligence model to diagnose COVID-19 based on chest CT images. (Methods) First, the two-dimensional fractional Fourier entropy was used to extract features. Second, a custom deep stacked sparse autoencoder (DSSAE) model was created to serve as the classifier. Third, an improved multiple-way data augmentation was proposed to resist overfitting. (Results) Our DSSAE model obtains a micro-averaged F1 score of 92.32% in handling a four-class problem (COVID-19, community-acquired pneumonia, secondary pulmonary tuberculosis, and healthy control). (Conclusion) Our method outperforms 10 state-of-the-art approaches.",

keywords = "COVID-19, Deep learning, autoencoder, fractional Fourier entropy",

author = "Wang, {Shui Hua} and Xin Zhang and Zhang, {Yu Dong}",

note = "Publisher Copyright: {\textcopyright} 2021 Copyright held by the owner/author(s). Publication rights licensed to ACM.",

year = "2021",

month = oct,

day = "5",

doi = "10.1145/3451357",

language = "English",

volume = "13",

journal = "ACM Transactions on Management Information Systems",

issn = "2158-656X",

number = "1",

}

TY - JOUR

T1 - DSSAE

T2 - Deep Stacked Sparse Autoencoder Analytical Model for COVID-19 Diagnosis by Fractional Fourier Entropy

AU - Wang, Shui Hua

AU - Zhang, Xin

AU - Zhang, Yu Dong

PY - 2021/10/5

Y1 - 2021/10/5

N2 - (Aim) COVID-19 has caused more than 2.28 million deaths till 4/Feb/2021 while it is still spreading across the world. This study proposed a novel artificial intelligence model to diagnose COVID-19 based on chest CT images. (Methods) First, the two-dimensional fractional Fourier entropy was used to extract features. Second, a custom deep stacked sparse autoencoder (DSSAE) model was created to serve as the classifier. Third, an improved multiple-way data augmentation was proposed to resist overfitting. (Results) Our DSSAE model obtains a micro-averaged F1 score of 92.32% in handling a four-class problem (COVID-19, community-acquired pneumonia, secondary pulmonary tuberculosis, and healthy control). (Conclusion) Our method outperforms 10 state-of-the-art approaches.

AB - (Aim) COVID-19 has caused more than 2.28 million deaths till 4/Feb/2021 while it is still spreading across the world. This study proposed a novel artificial intelligence model to diagnose COVID-19 based on chest CT images. (Methods) First, the two-dimensional fractional Fourier entropy was used to extract features. Second, a custom deep stacked sparse autoencoder (DSSAE) model was created to serve as the classifier. Third, an improved multiple-way data augmentation was proposed to resist overfitting. (Results) Our DSSAE model obtains a micro-averaged F1 score of 92.32% in handling a four-class problem (COVID-19, community-acquired pneumonia, secondary pulmonary tuberculosis, and healthy control). (Conclusion) Our method outperforms 10 state-of-the-art approaches.

KW - COVID-19

KW - Deep learning

KW - autoencoder

KW - fractional Fourier entropy

UR - http://www.scopus.com/inward/record.url?scp=85159196557&partnerID=8YFLogxK

U2 - 10.1145/3451357

DO - 10.1145/3451357

M3 - Article

AN - SCOPUS:85159196557

SN - 2158-656X

VL - 13

JO - ACM Transactions on Management Information Systems

JF - ACM Transactions on Management Information Systems

IS - 1

M1 - 2

ER -

DSSAE: Deep Stacked Sparse Autoencoder Analytical Model for COVID-19 Diagnosis by Fractional Fourier Entropy

Abstract

Keywords

UN SDGs

Access to Document

Other files and links

Fingerprint

Cite this