Wavelet and deep learning-based detection of SARS-nCoV from thoracic X-ray images for rapid and efficient testing

作者：

Highlights：

• Wavelet and AI-enabled testing protocol was developed for patients with SARS-nCoV.

• Xception-I trained and tested against 8000 datasets, 1552 of which were COVID-19.

• Wavelet-based deep CNN were validated using 5 and 10fold CV technique.

• Two-level Sym7 exhibited the test accuracy of 98.87%, followed by Bior2.6, 98.73%

• Precision, recall rate, and DSC for COVID-19 are 98%, 98%, and 99% respectively.

摘要

•Wavelet and AI-enabled testing protocol was developed for patients with SARS-nCoV.•Xception-I trained and tested against 8000 datasets, 1552 of which were COVID-19.•Wavelet-based deep CNN were validated using 5 and 10fold CV technique.•Two-level Sym7 exhibited the test accuracy of 98.87%, followed by Bior2.6, 98.73%•Precision, recall rate, and DSC for COVID-19 are 98%, 98%, and 99% respectively.

论文关键词：Medical imaging,Wavelets,rRT-PCR,COVID-19,Transfer learning

论文评审过程：Received 21 February 2021, Revised 2 June 2021, Accepted 20 July 2021, Available online 2 August 2021, Version of Record 5 August 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.115650