It’s difficult to fight a war when you don’t know how many troops the enemy has. In a pandemic, testing helps leaders understand the scale of the disease, and the severity of the measures they need to put in place.
Ideally, tests should be accurate and widely available. But experts are worried about the reliability and the rate of false negatives in Covid-19 tests, reported Forbes. Amazon Web Services (AWS) is supporting the work of researchers looking for more effective diagnostic methods.
From pooling data so leaders can make better decisions to boosting research, here are three ways AWS is lending its computing power, resources and tech expertise to the global fight against the pandemic.
1. Improve diagnosis
The University of British Columbia in Canada, along with Vancouver General Hospital and Vancouver Coastal Health Research Institute, is working on a way to better predict the presence of Covid-19 from CT scans. Using an open source AI model that looks at the pattern of white spots in a lung scan, doctors can determine if a patient has the coronavirus. They could even draw links between the images to the severity of the infection. This solves the availability and accuracy problems that come with test kits.
This method was developed in the Cloud Innovation Centre at the University of British Columbia, powered by AWS. The Centre provides students, staff and faculty with AWS cloud technology to advance research projects. It plans to present the model to the public, enabling healthcare professionals to submit CT scans for recommendations on probability of the presence of Covid-19.
The project was part of the global Diagnostic Development Initiative, which AWS started in March this year to speed up the development of accurate diagnostic tests. It is offering US$20 million in cloud credits and technical support, so researchers can use AWS computing power to process large amounts of data quickly and accelerate their work. “Given the need, the emphasis initially will be on Covid-19, but we will also consider other infectious disease diagnostic projects,” an Amazon blog stated.
2. Open data for better decisions
Researchers, healthcare leaders and policymakers rely on good data to understand the disease better, and make their decisions. Open data can help.
AWS Data Exchange pools information on Covid-related public records, from foot traffic from businesses to economic activity data and more. This helps decision-makers easily access available information on the coronavirus. The portal holds data from more than 2,300 sources.
The AWS COVID-19 data lake contains up-to-date and curated information on the spread and characteristics of Covid-19. Teams can run analyses with this data without having to waste time finding and extracting relevant datasets from multiple sources.
Researchers from the Chan Zuckerberg Biohub, a US-based medical science research centre used data from the AWS COVID-19 data lake to understand how the disease spreads. Business intelligence firm Domo included data from this portal in its coronavirus tracker for America.
The University of Swinburne Data for Social Good Cloud Innovation Centre built the BeatCovid19Now symptom tracker that helps individuals to monitor their, or their loved ones’, health. The tracker also informs public health decision makers on the current status and future developments of Covid-19.
BeatCovid19Now asks users questions about their symptoms. With enough users, local health authorities can know where and when the disease is spreading, and plan to protect the community.
This idea first materialised in a hackathon in March 2020. After the hackathon. AWS Partner Network (APN) Partner, Arq Group, completed the production build and hosting on AWS.
3. Support research
AWS has created tools to help speed up research on Covid-19. One of these is the CORD-19 Search. It uses machine learning to process a natural language question, combs through the tens of thousands of documents in its database, and provides precise answers along with the relevant references.
The search engine saw more than 5,500 unique searches from across 76 countries within just seven weeks of its launch, AWS wrote on its public sector blog. The volume of information collected on the coronavirus grows exponentially each day. It can be difficult to navigate, but CORD-19 Search can serve as a one-stop, integrated data repository for researchers.
AWS also contributed to the COVID-19 High Performance Computing Consortium. This platform consolidates computing resources from academics, industry and US-based federal government agencies for Covid-19 researchers.
The Consortium is currently supporting 81 research projects. One of these is a study to find a drug that can bind to and neutralise the coronavirus disease. The machine learning tools available through the Consortium can help to predict how to synthesise these suggested compounds. This can shorten the time to determine optimal ways to make these compounds from weeks to days.
AWS has also consolidated a centralised repository of all strains of the Covid-19 virus submitted to the National Centre for Biotechnology Information (NCBI) in the US. Researchers can access this genome sequence dataset, which is updated hourly, on NCBI’s website.
As the world continues to grapple with Covid-19, it will need all the information and innovation we have to offer. AWS will offer its resources and tech to help in strengthening global defenses against the virus.