A new sensor developed by researchers at Hopkins can detect communicable diseases like COVID-19, H1N1 and the Zika virus in saliva more accurately than traditional rapid tests at about the same speed.
The sensor relies on a combination of surface-enhanced Raman spectroscopy (SERS), machine learning, and large-area nanoimprint lithography. Researchers believe the technology could potentially boost public health safety measures in crowded locations.
The project began about two years ago, near the start of the COVID-19 pandemic. The researchers started with detecting SARS-CoV-2 as their primary goal, but their work eventually expanded to include other infectious diseases like Zika, H1N1 and the Marburg virus. The results were published earlier this spring in Nano Letters.
Ishan Barman is an associate professor of mechanical engineering with joint appointments in the Sidney Kimmel Comprehensive Cancer Center and the Russell H. Morgan Department of Radiology and Radiological Science. Barman is one of the senior authors of the paper and the principal investigator of the lab that created the sensor.
He discussed the project’s beginnings in an interview with The News-Letter.
“When the first wave [of COVID-19] hit, it was explosive enough; it was a problem of large enough significance that even if it hadn’t continued for as long as it has, it would still have been a problem worth solving,” he said. “At the end of the day, a crucial step in controlling outbreaks is the timely and accurate calculation of emerging viruses.”
Debadrita Paria is a postdoctoral fellow in the Barman Lab. In an email to The News-Letter, Paria noted that the process of making the sensor was complicated by pandemic restrictions.
“It required rigorous planning and execution. We used to have several Zoom meetings where several ideas would come up and then we would go to the lab and try to implement those and see what worked. Since we were working in shifts, we had to do the experiments in a limited timeframe,” Paria wrote.
While PCR and rapid antigen tests are currently used for SARS-CoV-2 detection, researchers have pointed out their limitations. PCR tests require intense sample processing, including the use of fluorescent markers, to detect if COVID-19 RNA is present in a sample. Rapid antigen tests lack accuracy and have been hard to find because of high demand. Another challenge, according to the paper published in Nano Letters, is storing and transporting samples to be processed.
Barman described how the sensor aims to counter those drawbacks.
“We were always thinking that we need better sensing technology that combines the salient features of what we know: RT-PCR, which has incredible sensitivity and specificity, with the convenience and speed of the rapid antigen test,” he said.
The new sensor relies on a saliva sample instead of a more invasive nasal swab. Additionally, its accuracy for detecting COVID-19 is around 92%, which is comparable to PCR, the current gold standard. Finally, the sensor can work quickly, giving a positive or negative test result in about 12 minutes, according to Barman. The lab plans to continue working on reducing that time.
Additionally, the sensor has built-in flexibility for SARS-CoV-2 mutations, meaning it will still be able to identify new variants. The team’s next goals are to work on identifying and differentiating these variants and to test real patient samples with the sensor to gauge how well it operates.
There are three components to how the sensor works: nanoimprint lithography, SERS and machine learning. The nanoimprint lithography provides a flexible surface for the saliva sample, using a field enhancing metal insulator antenna array to amplify the signal for the spectroscopy.
SERS reads the sample relying on inelastic scattering of light to characterize how unique molecules vibrate. If COVID-19 or another infectious disease is present in the sample, there will be characteristic vibration patterns on the spectroscopy readout.
The sensor then utilizes machine learning to determine if new samples are positive or negative based on what previous positive spectroscopy readouts looked like. According to the paper, using machine learning allows greater sensitivity and specificity to help overcome the noise from other unwanted biological specimens in the saliva sample.
According to Barman, it can be placed on doorknobs, masks and other locations to help facilitate on-site rapid testing because of the flexibility of the surface used for the sample. He noted that the portable device to be used in those instances is about two shoeboxes tall.
Barman highlighted that the new viral sensor has potential to be used as mass-testing technology not only for COVID-19 but also for other pathogens such as influenza, Zika virus and the Marburg virus.
“We wanted to create a tool that would be better at managing outbreaks in the future. Thinking beyond the pandemic was always an objective,” he said.
Source: The Johns Hopkins News-Letter.
