Explainable Application of Artificial Intelligence in Medical Imaging Diagnosis
Download as PDF
DOI: 10.25236/icetmr.2024.032
Corresponding Author
Guan Meng
Abstract
The purpose of this article is to discuss the application of XAI (Explainable Artificial Intelligence) in medical image diagnosis, so as to improve the accuracy and efficiency of diagnosis and the trust of doctors in AI (Artificial Intelligence) aided diagnosis. By summarizing the theoretical basis and technical overview of XAI, this study analyzes the design principles and key technologies of XAI system in medical image diagnosis in detail, including model visualization, local and global interpretable methods, and discusses clinical application cases, such as early lung cancer screening and breast cancer screening. It is found that XAI technology can significantly improve the accuracy and efficiency of medical image diagnosis, provide transparent decision-making process and understandable explanation, and enhance doctors' trust in AI-assisted diagnosis. To sum up, this study believes that XAI technology has a broad application prospect in medical image diagnosis. In the future, we should strengthen interdisciplinary cooperation and promote the innovative application and development of XAI technology; At the same time, pay attention to data security and privacy protection, improve doctors' acceptance of new technology, and promote the wide application of XAI technology in medical imaging diagnosis.
Keywords
Explainable Artificial Intelligence; Medical imaging diagnosis; Trust degree