Scholarlink

AI4AR is an innovative approach that integrates structured reporting (SR), artificial intelligence (AI), and education to improve radiology in a value-driven system. The AI4AR platform provides a modular framework for AI-augmented radiology; one that combines open-source and custom modules for automated image analysis, SR, and education. It integrates tools such as XNAT for data management, OHIF Viewer for visualization, and AI models for automated processing. The platform serves as a comprehensive resource for the creation, deployment, and evaluation of structured report templates, as well as supporting high-quality data collection and storage. AI improves reporting by optimizing report generation, ensuring consistency, and facilitating AI model training with structured datasets. This article explores the integration of AI and SR, with a focus on innovative radiology education. As a clinical use case, AI4AR has been applied to prostate cancer diagnosis, incorporating SR based on PI-RADS 2.1. A dataset of over 800 cases, annotated by expert radiologists, has been developed to train AI models for lesion detection and anatomical segmentation. The platform also offers educational modules, which enable radiologists to practice SR with or without AI assistance. AI4AR represents a significant step toward the integration of AI in radiology, promoting standardization, improving clinical decision making, and fostering the adoption of AI through education and collaboration.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

AI4AR: Reporting, Education, and Artificial Intelligence in the Era of Value-Based Radiology

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Rafał Jóźwiak,
  • Jan Mycka,
  • Ihor Mykhalevych,
  • Michał Gonet,
  • Piotr Sobecki,
  • Tomasz Jaworski,
  • Krzysztof Tupikowski,
  • Joanna Dołowy,
  • Tomasz Lorenc,
  • Anna Zacharzewska-Gondek

摘要

AI4AR is an innovative approach that integrates structured reporting (SR), artificial intelligence (AI), and education to improve radiology in a value-driven system. The AI4AR platform provides a modular framework for AI-augmented radiology; one that combines open-source and custom modules for automated image analysis, SR, and education. It integrates tools such as XNAT for data management, OHIF Viewer for visualization, and AI models for automated processing. The platform serves as a comprehensive resource for the creation, deployment, and evaluation of structured report templates, as well as supporting high-quality data collection and storage. AI improves reporting by optimizing report generation, ensuring consistency, and facilitating AI model training with structured datasets. This article explores the integration of AI and SR, with a focus on innovative radiology education. As a clinical use case, AI4AR has been applied to prostate cancer diagnosis, incorporating SR based on PI-RADS 2.1. A dataset of over 800 cases, annotated by expert radiologists, has been developed to train AI models for lesion detection and anatomical segmentation. The platform also offers educational modules, which enable radiologists to practice SR with or without AI assistance. AI4AR represents a significant step toward the integration of AI in radiology, promoting standardization, improving clinical decision making, and fostering the adoption of AI through education and collaboration.