2020-09-02 07:47:39
Moritz Knolle (1 and 2), Georgios Kaissis (1 and 2 and 3 and 4), Friederike Jungmann (1), Sebastian Ziegelmayer (1), Daniel Sasse (1), Marcus Makowski (1), Daniel Rueckert (2 and 4), Rickmer Braren (1) ((1) Department of diagnostic and interventional Radiology, Technical University of Munich, Munich, Germany, (2) Institute for Artificial Intelligence and Data Science in Medicine and Healthcare, Technical University of Munich, Munich, Germany, (3) OpenMined Research, (4) Department of Computing, Imperial College London, London, United Kingdom)
Abstract
For artificial intelligence-based image analysis methods to reach clinical applicability, the development of high-performance algorithms is crucial. For example, existent segmentation algorithms based on natural images are neither efficient in their parameter use nor optimized for medical imaging. Here we present MoNet, a highly optimized neural-network-based pancreatic segmentation algorithm focused on achieving high performance by efficient multi-scale image feature utilization.
Abstract (translated)
URL
https://arxiv.org/abs/2009.00872
PDF
https://arxiv.org/pdf/2009.00872.pdf