Predicting fluorescent labels in label-free microscopy images with pix2pix and adaptive loss in Light My Cells challenge

TL;DR

This paper introduces a deep learning approach based on pix2pix with an adaptive loss for in silico fluorescence labeling in microscopy, reducing the need for invasive procedures and improving labeling efficiency.

Contribution

The authors develop a novel pix2pix-based method with adaptive loss for in silico labeling, handling partially labeled datasets and multiple input modalities in microscopy.

Findings

Achieved promising performance in fluorescence image prediction

Effective training strategies for different input modalities

Demonstrated potential to replace invasive fluorescence labeling

Abstract

Fluorescence labeling is the standard approach to reveal cellular structures and other subcellular constituents for microscopy images. However, this invasive procedure may perturb or even kill the cells and the procedure itself is highly time-consuming and complex. Recently, in silico labeling has emerged as a promising alternative, aiming to use machine learning models to directly predict the fluorescently labeled images from label-free microscopy. In this paper, we propose a deep learning-based in silico labeling method for the Light My Cells challenge. Built upon pix2pix, our proposed method can be trained using the partially labeled datasets with an adaptive loss. Moreover, we explore the effectiveness of several training strategies to handle different input modalities, such as training them together or separately. The results show that our method achieves promising performance for in silico labeling. Our code is available at https://github.com/MedICL-VU/LightMyCells.