STResNet: Covid-19 Detection by ResNet Transfer Learning and Stochastic Pooling

Wei Wang, Shui Hua Wang, Yu Dong Zhang*

*Corresponding author for this work

Research output: Chapter in Book or Report/Conference proceedingConference Proceedingpeer-review

1 Citation (Scopus)

Abstract

Since 2019, COVID-19 has been spreading globally with a very rapid rate of transmission, resulting in a large number of confirmed diagnoses and deaths. The main interdiction measure currently in use for COVID-19 is the isolation of the confirmed population. For this reason, an effective and rapid diagnostic approach is particularly important. In this paper, we propose a deep learning framework (STResNet) for diagnosing COVID-19 from chest CT image slices. The proposed framework uses a modified residual network with 50 network layers as a backbone to extract features from chest CT slices and a support vector machine to classify the extracted features. Experiments show that the proposed framework has excellent performance. In the experiment based on a chest CT slices dataset, STResNet achieved accuracy of 93.81% ± 1.02%, MCC of 87.64% ± 2.02%, FMI of 93.83% ± 0.99%, sensitivity of 94.03% ± 1.07%, precision of 93.64% ± 1.54%, F1-score of 93.83% ± 0.99%, and specificity of 93.59% ± 1.67%. These demonstrate the excellent performance of the proposed framework with well balance and stability.

Original languageEnglish
Title of host publicationMedical Imaging and Computer-Aided Diagnosis - Proceedings of 2022 International Conference on Medical Imaging and Computer-Aided Diagnosis MICAD 2022
EditorsRuidan Su, Yudong Zhang, Han Liu, Alejandro F Frangi
PublisherSpringer Science and Business Media Deutschland GmbH
Pages489-502
Number of pages14
ISBN (Print)9789811667749
DOIs
Publication statusPublished - 2023
Externally publishedYes
EventInternational Conference on Medical Imaging and Computer-Aided Diagnosis, MICAD 2022 - Leicester, United Kingdom
Duration: 20 Nov 202221 Nov 2022

Publication series

NameLecture Notes in Electrical Engineering
Volume810 LNEE
ISSN (Print)1876-1100
ISSN (Electronic)1876-1119

Conference

ConferenceInternational Conference on Medical Imaging and Computer-Aided Diagnosis, MICAD 2022
Country/TerritoryUnited Kingdom
CityLeicester
Period20/11/2221/11/22

Keywords

  • ResNet-50
  • Stochastic pooling
  • Support vector machine

Fingerprint

Dive into the research topics of 'STResNet: Covid-19 Detection by ResNet Transfer Learning and Stochastic Pooling'. Together they form a unique fingerprint.

Cite this