Health AI Central
Share:
Peer-Reviewed Publication
J Imaging2023;9(8)August 8, 2023Journal Article

Biased Deep Learning Methods in Detection of COVID-19 Using CT Images: A Challenge Mounted by Subject-Wise-Split ISFCT Dataset.

Shiva Parsarad1,2, Narges Saeedizadeh1,3, Ghazaleh Jamalipour Soufi4, Shamim Shafieyoon4, Farzaneh Hekmatnia5, Andrew Parviz Zarei5, Samira Soleimany4, Amir Yousefi4, Hengameh Nazari4, Pegah Torabi4, Abbas S Milani6, Seyed Ali Madani Tonekaboni7, Hossein Rabbani1, Ali Hekmatnia4, Rahele Kafieh1,8
1Medical Image and Signal Processing Research Center, School of Advanced Technologies in Medicine, Isfahan University of Medical Sciences, Isfahan JM76+5M3, Iran.
2Law, Economics, and Data Science Group, Department of Humanities, Social and Political Science, ETH Zurich, 8092 Zurich, Switzerland.
3Institute for Intelligent Systems Research and Innovation, Deakin University, Melbourne, VIC 3125, Australia.
4Department of Radiology, School of Medicine, Isfahan University of Medical Sciences, Isfahan JM76+5M3, Iran.
5St. George's Hospital, London SW17 0RE, UK.
6School of Engineering, University of British Columbia, Kelowna, BC V1V 1V7, Canada.
7Cyclica Inc., Toronto, ON M5J 1A7, Canada.
8Department of Engineering, Durham University, Durham DH1 3LE, UK.
PubMed PMC DOI PDF

Abstract

Accurate detection of respiratory system damage including COVID-19 is considered one of the crucial applications of deep learning (DL) models using CT images. However, the main shortcoming of the published works has been unreliable reported accuracy and the lack of repeatability with new datasets, mainly due to slice-wise splits of the data, creating dependency between training and test sets due t…

Create a free account to keep reading

Free members get 10 full research views every month across publications, clinical trials, FDA clearances, adverse events, and NIH grants. No credit card required.

Create Free AccountSign In

Want unlimited research access? See Pro plans

Data Accuracy Notice: Research intelligence on Health AI Central is aggregated from public sources (PubMed, ClinicalTrials.gov, FDA, NIH, CMS, and others) and refreshed nightly. Classifications and derived metrics are produced by automated methods described in our Methodology. We recommend verifying critical data points against the primary sources before making decisions.