Applications of the ESPNet architecture in medical imaging
Sachin Mehtaa, Nicholas Nuechterleina, Ezgi Mercanb c, Beibin Lia, Shima Nofallaha, Wenjun Wua, Ximing Lua, Anat Caspia, Mohammad Rastegaria, Joann G. Elmored, Hannaneh Hajishirzia, Linda G. Shapiroa
- aUniversity of Washington, Seattle, WA, United States
- bSeattle Children’s Hospital, Seattle, WA, United States
- cSeattle Children's Hospital, Seattle, WA, United States
- dUniversity of California, Los Angeles, CA, United States
Papers open inside SciRah from the original publisher or repository; SciRah never hosts them.
Abstract
medical image analysis [6,12,13] . To demonstrate the modeling power of the ESPNet architecture, we study the application of ESPNet to two different medical imaging tasks: (1) tissue-level segmentation of breast biopsy WSIs and (2) tumor segmentation in 3D brain MR images. Our results show that the ESPNet architecture learns meaningful representations efficiently that allows it to deliver good
TL;DRThis work studies the application of ESPNet to two different medical imaging tasks: tissue-level segmentation of breast biopsy WSIs and tumor segmentation in 3D brain MR images. (Semantic Scholar)