Checklist for responsible deep learning modeling of medical images based on COVID-19 detection studies

PATTERN RECOGNITION

卷:118

Article

Hryniewska, Weronika¹  Bombinski, Przemyslaw²  Szatkowski, Patryk²  Tomaszewska, Paulina¹  Przelaskowski, Artur¹  Biecek, Przemyslaw^1,3 

[1] Warsaw Univ Technol, Fac Math & Informat Sci, Warsaw, Poland

[2] Med Univ Warsaw, Dept Pediat Radiol, Warsaw, Poland

[3] Univ Warsaw, Fac Math Informat & Mech, Warsaw, Poland

关键词: COVID-19;    Lungs;    Computed tomography;    X-ray;    Explainable AI;    Deep learning;   

DOI  :  10.1016/j.patcog.2021.108035

来源: Elsevier

【 摘 要 】

The sudden outbreak and uncontrolled spread of COVID-19 disease is one of the most important global problems today. In a short period of time, it has led to the development of many deep neural network models for COVID-19 detection with modules for explainability. In this work, we carry out a systematic analysis of various aspects of proposed models. Our analysis revealed numerous mistakes made at different stages of data acquisition, model development, and explanation construction. In this work, we overview the approaches proposed in the surveyed Machine Learning articles and indicate typical errors emerging from the lack of deep understanding of the radiography domain. We present the perspective of both: experts in the field -radiologists and deep learning engineers dealing with model explanations. The final result is a proposed checklist with the minimum conditions to be met by a reliable COVID-19 diagnostic model. (c) 2021 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license ( http://creativecommons.org/licenses/by-nc-nd/4.0/ )

【 授权许可】

Free