On Generative Modeling of Cell Shape Using 3D GANs

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.