AI Device Enhances Covid Test Accuracy: Lancaster University

Lancaster

A new AI-assisted molecular diagnostic platform capable of identifying variants of COVID-19 and other infectious diseases has been developed by scientists in the UK.

Scientists from Lancaster University, the University of Surrey and Brunel University London, in collaboration with partners from the NHS, GB Electronics (UK) Ltd and Vidiia Ltd have created the platform known as VIDIIA Hunter (VH6).

The low cost, portable device uses reverse transcription loop-mediated isothermal amplification (RT-LAMP) technology in combination with an artificial intelligence (AI)-based deep learning model and could play a crucial role in preventing future pandemics due to its accuracy and versatility.

Professor Muhammad Munir, Professor of Virology and Viral Zoonosis at Lancaster University said: "Incorporation of LAMP technology with advanced modules of AI has empowered the earliest, reliable and economical detection of infections, including COVID-19, and holds potential for the detection of diseases in both humans and animals, making it a tool of significant medical importance."

The new AI model has been trained to read the results of tests identifying infectious diseases, including COVID-19, and removes users' interpretation and errors whilst improving accuracy.

Professor Roberto La Ragione, Professor of Veterinary Microbiology and Pathology at the University of Surrey, added: "Lateral flow tests are an efficient way of testing if you have COVID-19, however, there has always been a question mark over their accuracy which has only been heightened with the emerging number of variants now in circulation. As COVID-19 continues to evolve, we need to evolve with it and have highly accurate tests that can be readily used without the need for laboratory facilities."

To confirm the accuracy of VH6 scientists tested 150 COVI-19 positive clinical nasal swabs with a range of viral loads and 250 negative samples provided by NHS Berkshire, Surrey Pathology and Royal Lancaster Infirmary, Lancaster. The test was found to be highly accurate with a detection rate of 98 per cent and a specificity of 100 per cent. Additional testing found the device detected all the COVID-19 variants which have circulated in the UK since December 2020.

Dr Aurore Poirier, first and co-corresponding author of the study and Research Fellow B at the University of Surrey, said: "The VH6 diagnostics platform has been approved for COVID-19 testing in the UK, but also has to potential to diagnose current and emerging infectious disease and antimicrobial resistance. Its portability, rapidity, accuracy, and affordability allow for near patient testing, in all laboratory and healthcare settings, including low-resources ones. The VIDIIA Hunter therefore has the potential to help control future outbreaks."

To monitor and track the spread of COVID-19 and other infectious diseases, the test is connected to a smartphone app that allows an operator to manage and track the patients and samples. Results and graphs are displayed on the app in as little time as 20-30 minutes and automatically connects to a cloud. The platform allows near-patient testing and has the potential to detect other infectious diseases such as tuberculosis and dengue fever, and antimicrobial resistance.

Unusually, the test can be used for human and animal healthcare which is a crucial step in identifying any future zoonotic diseases which could spread between the two.

The VH6 has now been approved for medical use in the United Kingdom under the UK Health Security Agency's Medical Devices (Coronavirus Test Device Approvals, CTDA) Regulations 2022 and is CE-IVD marked and MHRA registered.

A study using this diagnostic platform has been published in Frontiers in Molecular Biosciences.

/Public Release. This material from the originating organization/author(s) might be of the point-in-time nature, and edited for clarity, style and length. Mirage.News does not take institutional positions or sides, and all views, positions, and conclusions expressed herein are solely those of the author(s).View in full here.