as soon as is submitted to ZB.
METGAN: Generative tumour inpainting and modality synthesis in light sheet microscopy.
In: (2022 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 4-8 January 2022, Waikoloa). 2022. 3230-3240 (Proceedings - 2022 IEEE/CVF Winter Conference on Applications of Computer Vision, WACV 2022)
Novel multimodal imaging methods are capable of generating extensive, super high resolution datasets for preclinical research. Yet, a massive lack of annotations prevents the broad use of deep learning to analyze such data. In this paper, we introduce a novel generative method which leverages real anatomical information to generate realistic image-label pairs of tumours. We construct a dualpathway generator, for the anatomical image and label, trained in a cycle-consistent setup, constrained by an independent, pretrained segmentor. Our method performs two concurrent tasks: domain adaptation and semantic synthesis, which, to our knowledge, has not been done before. The generated images yield significant quantitative improvement compared to existing methods that specialize in either of these tasks. To validate the quality of synthesis, we train segmentation networks on a dataset augmented with the synthetic data, substantially improving the segmentation over the baseline.
Scopus
Cited By
Cited By
Altmetric
1
Annotations
Special Publikation
Hide on homepage
Publication type
Article: Conference contribution
Keywords
Autoencoders ; Gans ; Grouping And Shape ; Medical Imaging/imaging For Bioinformatics/biological And Cell Microscopy Deep Learning ; Neural Generative Models ; Segmentation
Language
english
Publication Year
2022
HGF-reported in Year
2022
Conference Title
2022 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV)
Conference Date
4-8 January 2022
Conference Location
Waikoloa
Quellenangaben
Pages: 3230-3240
Institute(s)
Institute for Tissue Engineering and Regenerative Medicine (ITERM)
POF-Topic(s)
30205 - Bioengineering and Digital Health
Research field(s)
Enabling and Novel Technologies
PSP Element(s)
G-505800-001
Scopus ID
85126091417
Erfassungsdatum
2022-07-14