Director DayOne Initiative, Senior Project Manager ICT, Public Relations
Genedata simplifies cell analysis
Basel – The Basel-based company Genedata has launched a new solution based on artificial learning to facilitate the identification of substances that can alter a cell’s phenotype.
Genedata’s new image analysis was developed in the context of its Early Access Program to Deep Learning for High Content Screening (HCS), which identifies substances such as molecules that can alter a cell’s phenotype.
HCS is used in the development of new medications, and the new image analysis is capable of automatically analysing a high volume of microscopy images, explains Genedata in a statement. The new software will enable pharmaceutical R&D organizations to more than halve the amount of time it takes to accurately identify HCS images and detect new phenotypes.
The Basel-based company is pioneering the use of deep learning technology, a type of artificial intelligence that uses convolutional neural networks (CNNs).
“With deep learning-based CNNs, we have been able to reduce the classification time by a large margin compared to classical methods while obtaining results quality equal to human experts,” explained Oliver Dürr, a deep learning expert from the Zurich University of Applied Sciences (ZHAW).
The algorithm used by Genedata only searches for what separates images from each other, which enables the discovery of new phenotypes. This approach not only slashes the time required for image analysis, it also empowers scientists not trained to develop classical image analysis protocols to perform HCS, thus further reducing image development time and lowering costs.