Jump to content

AI-powered High Content Screening Analytics with Genedata Imagence 3.0

Presented at SLAS2021 Digital

In this tutorial we showcase the future of automated HCS image analysis in Biopharma R&D, Genedata Imagence. Imagence combines state-of-the-art deep learning methods with protocols for efficient training and incremental training data enrichment, features intuitive representations of data and knowledge, and runs on a fully scalable infrastructure. Matching the needs of today’s most challenging bioassays, it allows screening scientists to reliably detect stable endpoints for primary drug response, assess toxicity and safety-relevant effects, and to discover new phenotypes and compound classes. 

In this tutorial we show how Imagence enables training of a new deep learning model in a few simple steps and how to apply the trained model to process a production HCS, producing results on an industrial scale with superior quality. We also present novelties in Imagence version 3: The incremental training data enrichment protocol which enables efficient retraining of a network to automatically adapt to slightly altered experimental conditions and a new classification uncertainty measure that allows to rapidly spot any data quality issues.  These are examples of Imagence’s continuous development towards bringing ultimate efficiency and ease-of-use to HCS analysis.

Request Resource

By submitting my data, I give consent to the collection, processing and use of my personal data in accordance with the Genedata privacy policy