A deep learning-based workflow for high-throughput and high-quality widefield fluorescent imaging of 3D samples

Edvin Forsgren, Christoffer Edlund, Miniver Oliver, Kalpana Barnes, Timothy R Jackson, Rickard Sjögren

Keywords: Convolutional Neural Network, Fluorescent 3D cell imaging, Conditional Generative Adversarial Network, Fluorescent live-cell imaging, High-throughput, Z-sweep, OSA-blocks, Tumor Spheroids

TL;DR: By combining faster z-sweep acquistions of 3D tumor spheroids with deep learning techniques we can increase the throughput of high-quality widefield fluorescent images by up to 100 times.

Abstract: 3D widefield fluorescent microscopic (wFLM) imaging is a widespread technology used to study living three-dimensional samples, such as tumor spheroids during drug development. However, 3D wFLM imaging suffers from a severe trade-off between image quality and throughput limiting its applicability. In this project, we present a novel workflow that enables high-throughput 3D wFLM imaging, which has previously been impossible, and apply it to fluorescent indicators of cell health within 3D tumor spheroids. The workflow combines deep learning with state-of-the-art live-cell imaging techniques to speed up the acquisition of a fluorescent image of a three-dimensional sample by a factor of a hundred.

Paper Type: methodological development

Primary Subject Area: Image Acquisition and Reconstruction

Secondary Subject Area: Image Synthesis

Paper Status: original work, not submitted yet

Source Code Url: Will be made available later.

Data Set Url: Not available atm. due to corporate reasons. Might be available at a later stage.

Registration: I acknowledge that publication of this at MIDL and in the proceedings requires at least one of the authors to register and present the work during the conference.

Authorship: I confirm that I am the author of this work and that it has not been submitted to another publication before.