Jun 3, 2022
0 0

New method for highly accurate and sensitive virus identification

New method for highly accurate and sensitive virus identification

A highly accurate and sensitive virus identification technique using Raman spectroscopy, a handheld virus capture device and machine learning could enable real-time virus detection and identification to help fight future pandemics, according to a Pennsylvania State University-led team of researchers.

“This virus detection method is label-free and does not target any specific virus, allowing us to identify potential new strains of viruses,” says Shanxi Huang, assistant professor of electrical and biomedical engineering and co-author of the study, published today (June 2) in Proceedings of the National Academy of Sciences. It is also fast, so it is suitable for rapid screening in crowded public places. “In addition, the rich Raman characteristics, together with machine learning analysis, allow a deeper understanding of the virus structures.”

Raman spectroscopy detects unique vibrations in molecules by detecting shifts when a beam of laser light causes these vibrations. Virus trapping will use a tool known as a microfluidic device to trap viruses between forests of aligned carbon nanotubes.

Microfluidic devices use very small volumes of body fluids on a microchip for medical and laboratory research. Such a device could use virus cultures, saliva, nasal swabs, or even exhaled breath, including samples collected on site during an outbreak. The carbon nanotube scaffolding will filter out any foreign matter or background molecules from the carrier or ambient air that can make accurate readings difficult.

“The fact that we use carbon nanotubes to enrich samples is very useful because in this way we enrich the sample with viruses and eliminate other bionoises that are undesirable when looking for a virus,” said Mauricio Terrones, Evan Pugh University Professor, Verna M Professor of Physics. Villamana and co-author of the study.

Once the samples are taken and the Raman microscope examines them, the machine learning aspect comes into play. The researchers collected Raman spectra from three different categories of viruses: human respiratory viruses, avian viruses, and enteroviruses. This data is then used to train a machine learning model, a convolutional neural network that identifies viruses.

“Once the machine learning model has been trained, given the unknown Raman spectrum of an unknown virus, our machine learning model can automatically recognize the type of virus,” says Sharon Huang, associate professor of information science and technology and the study’s corresponding author. “This includes, for example, recognition of the type of influenza, whether it is influenza A or influenza B, and the model can even recognize virus subtypes such as H1N1 or H3N2.”

According to the researchers, the benefits of such a device are numerous, especially in a rapidly developing outbreak.

“By providing a fast and label-free virus detection device, this approach will allow healthcare professionals to more closely track the evolution of the virus,” said Yin-Ting Yeh, assistant research professor at the Eberly College of Science and co-author of the study.

“Although the use of machine learning for Raman signal processing is not new in itself,” says Elodie Gedin, senior researcher, Division of Systems Genomics, NIH and co-author of the study. What makes this approach novel is the combination of a portable virus capture device, collection of Raman spectra from captured viruses on the device, and fast and accurate classification of viruses using a machine learning model.” This approach to real-time virus detection is especially timely. to combat current and future outbreaks”

Article Categories:

Leave a Reply