Johns Hopkins University in the United States has developed a new coronavirus sensor that can improve both accuracy and detection speed, and is expected to completely change the way the virus is detected. The new study, published in Nano Letters on the 29th, describes the new sensor, which does not require sample preparation and manipulation expertise and has strong advantages over existing detection methods, especially for large-scale population detection.

“The technology is as simple as putting a drop of saliva on the device and getting a negative or positive result,” said Ethan Barman, associate professor of mechanical engineering at Johns Hopkins University. The novelty is that it is a Label-free technology, which means no additional chemical modifications such as molecular labels or antibody functionalization are required. The sensor could eventually be used in wearable devices.

Barman said the new technology, which is not yet on the market, makes up for the limitations of the two most widely used methods of testing for the new coronavirus. PCR (polymerase chain reaction) tests are very accurate, but require complex sample preparation and take hours or even days to process results in the lab; another antigen test is less successful at detecting early-stage infections and asymptomatic cases, which may also lead to erroneous results.

The new sensor is almost as sensitive as PCR testing and as convenient as rapid antigen testing. During initial testing, the sensor demonstrated 92 percent accuracy in detecting SARS-CoV-2 in saliva samples, on par with PCR testing. The sensor has also been very successful in quickly identifying other viruses, including H1N1 and Zika.

Based on large-area nanoimprint lithography, surface-enhanced Raman spectroscopy, and machine learning techniques, the sensor can be tested at scale on rigid or flexible surfaces in a disposable chip format.

The key to the technology is a large-area, flexible field-enhanced metal-insulator antenna (FEMIA) array developed by the researchers. A saliva sample was placed on the material and analyzed using surface-enhanced Raman spectroscopy, which uses a laser to examine how the sample molecules vibrate. Since the nanostructured FEMIA significantly enhances the Raman signal of the virus, the system can rapidly detect the presence of the virus, even if only a small trace is present in the sample. Another major innovation in the system is the use of advanced machine learning algorithms to detect very subtle features in spectral data, allowing researchers to pinpoint the presence and concentration of the virus.

“Our platform goes beyond current testing for COVID-19,” Barman said. “We can use it for a broad range of tests for different viruses, for example, distinguishing 2019-nCoV from H1N1, or even variants. This is the main problem that the current rapid tests cannot easily solve.”