Paper Type

ERF

Abstract

This paper presents a systematic review of explainable artificial intelligence (XAI) methods applied to COVID-19 image analysis. We examined 154 records published between 2020 and 2024, identifying 45 distinct XAI techniques. Our analysis shows that a limited number of methods, primarily Grad-CAM, LIME, and SHAP, were frequently employed. While these techniques improve transparency in deep learning models, most studies rely on a single approach rather than integrating multiple methods to offer comprehensive interpretability. The findings highlight significant progress in making Deep Learning models more understandable for clinical decision-making, while also emphasizing the need for further research to combine and refine these techniques. This work aims to guide future efforts in developing transparent models for medical imaging applications.

Paper Number

2254

Author Connect URL

https://authorconnect.aisnet.org/conferences/AMCIS2025/papers/2254

Comments

SIGODIS

Author Connect Link

Share

COinS
 
Aug 15th, 12:00 AM

Explainable AI Methods in Medical Image Analysis

This paper presents a systematic review of explainable artificial intelligence (XAI) methods applied to COVID-19 image analysis. We examined 154 records published between 2020 and 2024, identifying 45 distinct XAI techniques. Our analysis shows that a limited number of methods, primarily Grad-CAM, LIME, and SHAP, were frequently employed. While these techniques improve transparency in deep learning models, most studies rely on a single approach rather than integrating multiple methods to offer comprehensive interpretability. The findings highlight significant progress in making Deep Learning models more understandable for clinical decision-making, while also emphasizing the need for further research to combine and refine these techniques. This work aims to guide future efforts in developing transparent models for medical imaging applications.

When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.