Litcius/Paper detail

Synthetic whole-slide image tile generation with gene expression profile-infused deep generative models

Francisco Carrillo‐Pérez, Marija Pizurica, Michael G. Ozawa, Hannes Vogel, Robert B. West, Christina S. Kong, Luis Javier Herrera, Jeanne Shen, Olivier Gevaert

2023Cell Reports Methods16 citationsDOIOpen Access PDF

Abstract

In this work, we propose an approach to generate whole-slide image (WSI) tiles by using deep generative models infused with matched gene expression profiles. First, we train a variational autoencoder (VAE) that learns a latent, lower-dimensional representation of multi-tissue gene expression profiles. Then, we use this representation to infuse generative adversarial networks (GANs) that generate lung and brain cortex tissue tiles, resulting in a new model that we call RNA-GAN. Tiles generated by RNA-GAN were preferred by expert pathologists compared with tiles generated using traditional GANs, and in addition, RNA-GAN needs fewer training epochs to generate high-quality tiles. Finally, RNA-GAN was able to generalize to gene expression profiles outside of the training set, showing imputation capabilities. A web-based quiz is available for users to play a game distinguishing real and synthetic tiles: https://rna-gan.stanford.edu/, and the code for RNA-GAN is available here: https://github.com/gevaertlab/RNA-GAN.

Topics & Concepts

AutoencoderRNAComputer scienceGene expressionTileArtificial intelligenceCode (set theory)Expression (computer science)GeneGenerative grammarGenerative modelSource codeGenerative adversarial networkComputational biologyRepresentation (politics)Deep learningSet (abstract data type)Pattern recognition (psychology)Computer visionBiologyGeneticsProgramming languagePoliticsPolitical scienceVisual artsArtLawGenerative Adversarial Networks and Image SynthesisAI in cancer detectionCell Image Analysis Techniques
Synthetic whole-slide image tile generation with gene expression profile-infused deep generative models | Litcius