AI 中文总结
该研究提出CheXtriev框架，采用图变换器提取胸部X光片特定解剖区域特征，其检索准确率与排名质量优于现有最优方法，尤其适用于少见病变。
AI 中文摘要
我们提出CheXtriev，这是一种基于图的、感知解剖结构的胸部X光片检索框架。与以往专注于全局特征的方法不同，我们的方法利用图变换器从特定解剖区域提取有用特征，还捕捉空间上下文以及解剖位置与病变之间的相互作用。这种基于循证解剖学的上下文化，形成了更丰富的感知解剖结构的表示，实现了更准确、有效且高效的检索，尤其是针对较少见的病变。CheXtriev在检索准确率上比现有最优的全局和局部方法高出18%至26%，在排名质量上高出11%至23%。代码可在该https URL获取。
英文摘要
We present CheXtriev, a graph-based, anatomy-aware framework for chest radiograph retrieval. Unlike prior methods focussed on global features, our method leverages graph transformers to extract informative features from specific anatomical regions. Furthermore, it captures spatial context and the interplay between anatomical location and findings. This contextualization, grounded in evidence-based anatomy, results in a richer anatomy-aware representation and leads to more accurate, effective and efficient retrieval, particularly for less prevalent findings. CheXtriv outperforms state-of-the-art global and local approaches by 18% to 26% in retrieval accuracy and 11% to 23% in ranking quality. The code is available at https://github.com/cvit-mip/chextriev.
CommentsAccepted at the 27th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI 2024)
