肺结节影像人工智能技术现状与思考

Current status and thinking on artificial intelligence technology in pulmonary nodule imaging

我国是肺癌高发国家,面对影像科医师匮乏与患病人数日益增长的医疗需求问题,肺结节人工智能技术的应用有望有效地平衡当前的医疗资源。本文旨在陈述肺结节影像人工智能技术发展的现状,探讨如何正确看待肺结节人工智能技术及其核心要素和目前面临的挑战等问题。指出肺结节人工智能技术已经初见成效,跨学科的深度合作、规范标注的影像大数据训练集及适合临床场景的多任务模型是其未来发展的核心推动力。

Lung cancer has high incidence in China. Faced with the lack of imaging physicians and the growing medical needs of large patient populations, the application of artificial intelligence technology in pulmonary nodules imaging can effectively balance current medical resources. This paper expounds the current situation of the development of artificial intelligence technology in pulmonary nodule imaging, discusses how to correctly treat this techonology the core elements and the faced challenges. It is pointed out that the artificial intelligence technology in lung nodule imaging has achieved initial success. The standardized image data training set and the multi-task model suitable for clinical scenes are the core elements of future product landing.