Checklist for responsible deep learning modeling of medical images based on COVID-19 detection studies
This work addresses the problem of unreliable deep learning models for medical image analysis, specifically for COVID-19 detection, by providing a checklist for researchers and developers to improve model trustworthiness and clinical utility.
This paper systematically analyzes deep neural network models for COVID-19 detection, identifying numerous mistakes in data acquisition, model development, and explanation construction. The authors propose a checklist of minimum conditions for reliable COVID-19 diagnostic models, based on perspectives from both radiologists and deep learning engineers.
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.