Explainability, transparency and black box challenges of AI in radiology: impact on patient care in cardiovascular radiology

The Egyptian Journal of Radiology and Nuclear Medicine, Год журнала: 2024, Номер 55(1)

Опубликована: Сен. 13, 2024

Abstract The integration of artificial intelligence (AI) in cardiovascular imaging has revolutionized the field, offering significant advancements diagnostic accuracy and clinical efficiency. However, complexity opacity AI models, particularly those involving machine learning (ML) deep (DL), raise critical legal ethical concerns due to their "black box" nature. This manuscript addresses these by providing a comprehensive review technologies imaging, focusing on challenges implications black box phenomenon. We begin outlining foundational concepts AI, including ML DL, applications imaging. delves into issue, highlighting difficulty understanding explaining decision-making processes. lack transparency poses for acceptance deployment. discussion then extends AI's opacity. need explicable systems is underscored, with an emphasis principles beneficence non-maleficence. explores potential solutions such as explainable (XAI) techniques, which aim provide insights without sacrificing performance. Moreover, impact explainability patient outcomes examined. argues development hybrid models that combine interpretability advanced capabilities systems. It also advocates enhanced education training programs healthcare professionals equip them necessary skills utilize effectively. Patient involvement informed consent are identified components deployment healthcare. Strategies improving engagement discussed, emphasizing importance transparent communication education. Finally, calls establishment standardized regulatory frameworks policies address unique posed By fostering interdisciplinary collaboration continuous monitoring, medical community can ensure responsible ultimately enhancing care outcomes.

Язык: Английский

The Role of Artificial Intelligence and Machine Learning in Cardiovascular Imaging and Diagnosis: Current Insights and Future Directions

Cureus, Год журнала: 2024, Номер unknown

Опубликована: Окт. 24, 2024

Cardiovascular diseases (CVDs) are the major cause of mortality worldwide, emphasizing critical need for timely and accurate diagnosis. Artificial intelligence (AI) machine learning (ML) have become revolutionary tools in healthcare system with significant potential cardiovascular diagnosis imaging. AI ML techniques, including supervised unsupervised learning, logistic regression, deep models, neural networks, convolutional networks (CNNs), significantly advanced Applications echocardiography include left right ventricular segmentation, ejection fraction measurement, wall motion analysis. hold substantial promise revolutionizing imaging, demonstrating improvements diagnostic accuracy efficiency. This narrative review aims to explore current applications, advantages, challenges, future pathways highlighting their impact on different imaging modalities integration into clinical practice.

Язык: Английский