eduzhai > Applied Sciences > Engineering >

Attention-Guided Generative Adversarial Network to Address Atypical Anatomy in Modality Transfer

  • Save

... pages left unread,continue reading

Document pages: 6 pages

Abstract: Recently, interest in MR-only treatment planning using synthetic CTs (synCTs)has grown rapidly in radiation therapy. However, developing class solutions formedical images that contain atypical anatomy remains a major limitation. Inthis paper, we propose a novel spatial attention-guided generative adversarialnetwork (attention-GAN) model to generate accurate synCTs using T1-weighted MRIimages as the input to address atypical anatomy. Experimental results onfifteen brain cancer patients show that attention-GAN outperformed existingsynCT models and achieved an average MAE of 85.22$ pm$12.08, 232.41$ pm$60.86,246.38$ pm$42.67 Hounsfield units between synCT and CT-SIM across the entirehead, bone and air regions, respectively. Qualitative analysis shows thatattention-GAN has the ability to use spatially focused areas to better handleoutliers, areas with complex anatomy or post-surgical regions, and thus offerstrong potential for supporting near real-time MR-only treatment planning.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...