{"type":"rich","version":"1.0","provider_name":"Transistor","provider_url":"https://transistor.fm","author_name":"萌喵读文献-肿瘤学","title":"今日肿瘤学最高分文献 - 2025-03-05","html":"<iframe width=\"100%\" height=\"180\" frameborder=\"no\" scrolling=\"no\" seamless src=\"https://share.transistor.fm/e/60052e10\"></iframe>","width":"100%","height":180,"duration":4,"description":"欢迎收听本周的医学科技快讯！最新研究发现，一种两阶段深度学习模型在CT图像上自动检测肾上腺结节的能力表现出色。这项回顾性研究在内部数据集和外部实践中进行了测试，结果显示模型对右肾上腺结节检测的AUC达到0.98，左肾上腺结节为0.93，均具有高度敏感性和特异性。这项技术不仅提高了结节检测的准确性，还可能改善偶发性肾上腺结节的处理，为放射学领域带来革命性的变革。不要错过这一前沿进展，深入了解如何利用AI提高诊断准确性，就在今天的播客中。","thumbnail_url":"https://img.transistorcdn.com/3WX36CzvFOrNjlF7WfWjzS4HdcMIMmzMs2bmPS-ejLU/rs:fill:0:0:1/w:400/h:400/q:60/mb:500000/aHR0cHM6Ly9pbWct/dXBsb2FkLXByb2R1/Y3Rpb24udHJhbnNp/c3Rvci5mbS83NDNm/MjQ1N2E5NjhjMzVj/NjM0ZTA3YzY2NDE2/ZmIyMi5qcGc.webp","thumbnail_width":300,"thumbnail_height":300}