eduzhai > Applied Sciences > Engineering >

Automated Intracranial Artery Labeling using a Graph Neural Network and Hierarchical Refinement

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 11 pages

Abstract: Automatically labeling intracranial arteries (ICA) with their anatomicalnames is beneficial for feature extraction and detailed analysis ofintracranial vascular structures. There are significant variations in the ICAdue to natural and pathological causes, making it challenging for automatedlabeling. However, the existing public dataset for evaluation of anatomicallabeling is limited. We construct a comprehensive dataset with 729 MagneticResonance Angiography scans and propose a Graph Neural Network (GNN) method tolabel arteries by classifying types of nodes and edges in an attributedrelational graph. In addition, a hierarchical refinement framework is developedfor further improving the GNN outputs to incorporate structural and relationalknowledge about the ICA. Our method achieved a node labeling accuracy of 97.5 ,and 63.8 of scans were correctly labeled for all Circle of Willis nodes, on atesting set of 105 scans with both healthy and diseased subjects. This is asignificant improvement over available state-of-the-art methods. Automaticartery labeling is promising to minimize manual effort in characterizing thecomplicated ICA networks and provides valuable information for theidentification of geometric risk factors of vascular disease. Our code anddataset are available at this https URL.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×