Paschali Magdalini, Chen Zhihong, Blankemeier Louis, Varma Maya, Youssef Alaa, Bluethgen Christian, Langlotz Curtis, Gatidis Sergios, Chaudhari Akshay
From the Stanford Center for Artificial Intelligence in Medicine and Imaging, 1701 Page Mill Rd, Palo Alto, CA 94304 (M.P., Z.C., L.B., M.V., A.Y., C.B., C.L., S.G., A.C.); Departments of Radiology (M.P., Z.C., A.Y., C.L., S.G., A.C.), Electrical Engineering (L.B.), Computer Science (M.V.), Medicine (C.L.), and Biomedical Data Science (C.L., A.C.), Stanford University, Stanford, Calif; and Department of Diagnostic and Interventional Radiology, University Hospital Zurich, University of Zurich, Zurich, Switzerland (C.B.).
Radiology. 2025 Feb;314(2):e240597. doi: 10.1148/radiol.240597.
Recent advances in artificial intelligence have witnessed the emergence of large-scale deep learning models capable of interpreting and generating both textual and imaging data. Such models, typically referred to as foundation models (FMs), are trained on extensive corpora of unlabeled data and demonstrate high performance across various tasks. FMs have recently received extensive attention from academic, industry, and regulatory bodies. Given the potentially transformative impact that FMs can have on the field of radiology, radiologists must be aware of potential pathways to train these radiology-specific FMs, including understanding both the benefits and challenges. Thus, this review aims to explain the fundamental concepts and terms of FMs in radiology, with a specific focus on the requirements of training data, model training paradigms, model capabilities, and evaluation strategies. Overall, the goal of this review is to unify technical advances and clinical needs for safe and responsible training of FMs in radiology to ultimately benefit patients, providers, and radiologists.
人工智能的最新进展见证了能够解释和生成文本及图像数据的大规模深度学习模型的出现。这类模型通常被称为基础模型（FMs），它们在大量未标记数据的语料库上进行训练，并在各种任务中表现出高性能。基础模型最近受到了学术界、行业和监管机构的广泛关注。鉴于基础模型可能对放射学领域产生变革性影响，放射科医生必须了解训练这些特定于放射学的基础模型的潜在途径，包括理解其益处和挑战。因此，本综述旨在解释放射学中基础模型的基本概念和术语，特别关注训练数据的要求、模型训练范式、模型能力和评估策略。总体而言，本综述的目标是将技术进步与临床需求统一起来，以便在放射学中安全、负责地训练基础模型，最终使患者、医疗服务提供者和放射科医生受益。
