On Generative Modeling of Cell Shape Using 3D GANs

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216224%3A14330%2F19%3A00107522" target="_blank" >RIV/00216224:14330/19:00107522 - isvavai.cz</a>

Výsledek na webu

<a href="http://dx.doi.org/10.1007/978-3-030-30645-8_61" target="_blank" >http://dx.doi.org/10.1007/978-3-030-30645-8_61</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1007/978-3-030-30645-8_61" target="_blank" >10.1007/978-3-030-30645-8_61</a>

Jazyk výsledku

angličtina

Název v původním jazyce

On Generative Modeling of Cell Shape Using 3D GANs

Popis výsledku v původním jazyce

The ongoing advancement of deep-learning generative models, showing great interest of the scientific community since the introduction of the generative adversarial networks (GAN), paved the way for generation of realistic data. The utilization of deep learning for the generation of realistic biomedical images allows one to alleviate the constraints of the parametric models, limited by the employed mathematical approximations. Building further upon the laid foundation, the 3D GAN added another dimension, allowing generation of fully 3D volumetric data. In this paper, we present an approach to generating fully 3D volumetric cell masks using GANs. Presented model is able to generate high-quality cell masks with variability matching the real data. Required modifications of the proposed model are presented along with the training dataset, based on 385 real cells captured using the fluorescence microscope. Furthermore, the statistical validation is also presented, allowing to quantitatively assess the quality of data generated by the proposed model.

Název v anglickém jazyce

On Generative Modeling of Cell Shape Using 3D GANs

Popis výsledku anglicky

The ongoing advancement of deep-learning generative models, showing great interest of the scientific community since the introduction of the generative adversarial networks (GAN), paved the way for generation of realistic data. The utilization of deep learning for the generation of realistic biomedical images allows one to alleviate the constraints of the parametric models, limited by the employed mathematical approximations. Building further upon the laid foundation, the 3D GAN added another dimension, allowing generation of fully 3D volumetric data. In this paper, we present an approach to generating fully 3D volumetric cell masks using GANs. Presented model is able to generate high-quality cell masks with variability matching the real data. Required modifications of the proposed model are presented along with the training dataset, based on 385 real cells captured using the fluorescence microscope. Furthermore, the statistical validation is also presented, allowing to quantitatively assess the quality of data generated by the proposed model.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/GA17-05048S" target="_blank" >GA17-05048S: Segmentace a trekování živých buněk v multimodálních obrazech</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2019

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Image Analysis and Processing – ICIAP 2019

ISBN

9783030306441

ISSN

0302-9743

e-ISSN

—

Počet stran výsledku

11

Strana od-do

672-682

Název nakladatele

Springer

Místo vydání

Trento

Místo konání akce

Trento, Italy

Datum konání akce

1. 1. 2019

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

000562008400059