Leveraging image complexity in macro-level neural network design for medical image segmentation

Tariq M Khan1, Syed S Naqvi2, Erik Meijering3

  • 1School of Computer Science and Engineering, University of New South Wales, Sydney, NSW, Australia. tariq045@egmail.com.

Scientific Reports
|December 24, 2022
PubMed
Summary

Image complexity guides medical image segmentation network choices. Median frequency best predicts optimal downsampling and network size (deep/shallow, large/lightweight) for performance on affordable hardware.

Related Concept Videos