Cell Cycle Phase Classification from Deep Learning-Predicted Images of Cell Organelles
Journal
Proceedings - IEEE 22nd International Conference on Bioinformatics and Bioengineering, BIBE 2022
ISBN
9781665484879
Date Issued
2022-01-01
Author(s)
Abstract
The cell cycle is the main process that regulates cell growth and development. The shape and size of cells and organelles change dynamically during the cell cycle. In this paper, deep neural networks (DNNs) were used to predict cell cycle phases from microscopic images. The fluorescent images of the nucleus were predicted from transmitted light image channels from a U-Net model to reduce fluorescent labeling and prevent phototoxicity. The predicted cell nucleus images, along with the transmitted light cell images and fluorescently labeled mitochondria images, were used to train the convolutional neural network ResNet34 to predict the cell cycle stage. Compared with only fluorescently Vabeled or transmitted light images, convolutional neural networks provide increased prediction accuracy with additional predicted nucleus images from transmitted light images, resulting in improved classification of cell cycle phases.
Subjects
cell cycle | confacal microscope | deep learning | image analysis
SDGs
Type
conference paper