Oxford Uni - Oct 202015 Mar 2021 21:08
Feeling good about BRH - Oxford Uni Oct 2020
Virus detection and identification in minutes using single-particle imaging and deep learning
Single-particle imaging combined with deep learning offers a promising alternative to traditional viral diagnostic methods, and has the potential for significant impact.
Our work demonstrates how single-particle fluorescence microscopy combined with deep learning can help to rapidly detect and classify viruses, including coronaviruses. Our approach of instantaneous labelling, rapid automated imaging, pre-processing and deep learning classifies viruses within minutes, avoiding the need for viral lysis or amplification and the associated cost, tedium and supply-chain issues.
The non-specific detection of intact viral particles (rather than genome fragments) can report directly on infectivity, and has the advantages of speed (results within 2-5 minutes), the ability to detect multiple virus types in a single labelling step, and robustness against potential mutations in the viral genome. Our algorithms are extremely versatile and can be trained to differentiate between many different viruses, independently of how they are labelled, immobilized and imaged. Given its simplicity and rapid nature, our technology could also be used outside of specialized laboratories, such as in airports, workplaces and care homes. These unique capabilities should enable extremely rapid, mobile, and real-time analysis of patient and community samples during pandemic situations.
https://www.medrxiv.org/content/10.1101/2020.10.13.20212035v1.full.pdf
Sign In
Follow the stocks
