Univ Labs Developing Covid-19, Outbreak Test Devices

Lena Dolgosheina holds vials with dyed RNA from the Mango system (Simon Fraser University)

20 Mar. 2020. Two university labs are creating tests or monitoring devices to detect novel coronavirus, or Covid-19 viruses, and spot other viral outbreaks when they occur. A rapid test for Covid-19 viruses is in development at Simon Fraser University in Burnaby, British Columbia, Canada, while a device for revealing flu-like illnesses in health care waiting rooms is a product of University of Massachusetts in Amherst.

Fast, simple Covid-19 test

Researchers from the lab of molecular biologist Peter Unrau at Simon Fraser University are applying their current work on visualizing RNA molecules to the task of quickly identifying the presence of Sars-CoV-2 viruses responsible for Covid-19 infections. Unrau and postdoctoral fellow Lena Dolgosheina developed a technology they call Mango that binds together RNA molecules and fluorescent dyes. The term Mango comes from the bright orange color that appears when a target RNA molecule is detected.

Mango uses aptamers, short strands of RNA that occur in nature, but can also be synthesized. Aptamers seek out and bind tightly to target RNA molecules, which makes them useful for biological sensing tasks. In a paper published on 9 March in the journal Nature Communications, Unrau and colleagues demonstrate Mango’s ability to detect and visualize RNA molecules in live cells, with high contrast and single-molecule sensitivity.

These properties make Mango useful for detection of viruses with characteristic RNA signatures, such as Sars-CoV-2. Another desirable quality is its speed. Unlike reverse transcription polymerase chain reaction, or RT-PCR, analysis used in most Covid-19 tests today, Mango does not requires expensive lab equipment or hours to complete. The SFU team says the Mango-based test can return results in minutes rather than hours.

The Canadian Institute of Health Research is funding Unrau’s lab with a $517,000 award to adapt Mango to a Covid-19 test. “We are made of molecules so when something goes wrong within a cell it happens at the molecular level,” says Unrau in a university statement. “We are using the Mango system as a catalyst, to allow us to not only extend fundamental research questions but also to detect pathogens like the coronavirus, faster and more efficiently.”

Spotting viral outbreaks

Researchers at UMass in Amherst are addressing a related problem in identifying pandemics, spotting the presence of viral illnesses where sick people congregate, such as waiting areas in health care facilities. A team from the university’s Mobile Sensing and Ubiquitous Computing, or Mosaic, lab is developing a system called FluSense that uses audio and visual signals, analyzed with machine learning to give early warning of flu or other viral outbreaks, before showing up in public health statistics.

FluSense is the creation of doctoral candidate Forsad Al Hossain, working with lab co-director Tauhidur Rahman, a professor of computer and information sciences. The portable system fits in a table-top box, with microphones and thermal cameras that form images from infrared signals rather than visible light. Al Hossain, Rahman, and colleagues describe FluSense in the March 2020 issue of the journal Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies (paid subscription required).

FluSense is powered by a Raspberry Pi processor, an open-source system that collects the audio and thermal imaging signals in waiting areas. The data are non-identified, and processed with edge computing, where data crunching occurs where collected — i.e., at the edge — rather than centrally in the cloud or elsewhere. The UMass team developed a machine-learning model to analyze the captured coughing sounds and thermal images inside FluSense, and provide an alert in almost real time that a flu outbreak may be occurring.

“I’ve been interested in non-speech body sounds for a long time,” says Rahman in a UMass statement. “I thought if we could capture coughing or sneezing sounds from public spaces where a lot of people naturally congregate, we could utilize this information as a new source of data for predicting epidemiologic trends.”

To prove the concept, the researchers used FluSense in four public waiting areas at the university’s health services, over a seven-month period in 2018-19. The team captured more than 21 million audio samples and 350,000 thermal images with FluSense, and generated predictions of patient counts that highly correlated (+0.95, where +1.0 is perfect correlation) with clinical records of flu cases during that period. The researchers next plan to expand FluSense testing to other environments.

Tauhidur Rahman and Forsad Al Hossain display their FluSense device. (University of Massachusetts in Amherst)

More from Science & Enterprise:

• FDA Clears High-Volume Covid-19 Test
• Skin Cancer Apps Shown Misleading, Unreliable
• A.I. Heart Ultrasound Software Cleared by FDA
• Incubator Seeks Heart, Lung, Blood Device Entrepreneurs
• Smartphone Saliva Test Detects Infectious Diseases

*     *     *