Transfer Learning in Optical Microscopy

Warning

This publication doesn't include Faculty of Arts. It includes Faculty of Informatics. Official publication website can be found on muni.cz.
Authors

KOZLOVSKÝ Martin WIESNER David SVOBODA David

Year of publication 2021
Type Article in Proceedings
Conference Simulation and Synthesis in Medical Imaging
MU Faculty or unit

Faculty of Informatics

Citation
Web https://2021.sashimi-workshop.org/
Doi http://dx.doi.org/10.1007/978-3-030-87592-3_8
Keywords Fluorescence microscopy; Phase-contrast microscopy; GAN; Image synthesis; Machine learning
Description Image synthesis is nowadays a very rapidly evolving branch of deep learning. One of possible applications of image synthesis is an image-to-image translation. There is currently a lot of focus orientated to applications of image translation in medicine, mainly involving translation between different screening techniques. One of other possible use of image translation in medicine and biology is in the task of translation between various imaging techniques in modern microscopy. In this paper, we propose a novel method based on DenseNet architecture and we compare it with Pix2Pix model in the task of translation from images imaged using phase-contrast technique to fluorescence images with focus on usability for cell segmentation.
Related projects:

You are running an old browser version. We recommend updating your browser to its latest version.