Intellegens wins Innovate UK grant to model Covid-19 data to help manage future pandemics
Intellegens, the Cambridge-based artificial intelligence company, has been given Innovate UK funding to model global Covid-19 data, with the aim of improving the management of future pandemics.
The Barclays Eagle Labs-based company has been granted £48,916 to build an interactive, predictive tool that can help governments, healthcare providers and companies predict the progression of pandemics.
Building on its Alchemite Engine and Analytics products, its tool will feature a dashboard that Intellegens says will accurately summarise predictions for the geographic distribution, timing, and patient characteristics of future outbreaks of Covid-19 or other global pandemics.
Users will be able to change parameters on the tool to model how policy choices might impact on outcomes, which could prove valuable in informing decision-making.
Gareth Conduit, chief technology officer at Intellegens, said: “The emergence of Covid-19 has exposed a weakness in the global strategy for managing pandemics, which can significantly impact on human health, daily life, and world economics.
“Throughout the pandemic, different countries have adopted different approaches, with varying degrees of success.
“Globally, the challenges faced have been distinct with many countries facing a shortage of critical supplies including personal protective equipment, drugs, and ventilators.
“The key to managing future pandemics lies in modeling Covid-19 data from multiple sources, to compare and contrast the experiences of different countries at a granular level, and learn from best practice.
“This means bringing together information, in different formats, which may be inconsistent and, in some cases, incomplete.
“Standard data modelling tools struggle with data that is sparse or noisy, but the Intellegens’ Alchemite technology is different. Working with the information available, Alchemite accurately predicts missing values and extrapolates correlations between all parameters - significantly reducing the amount of time and money spent on scrutinising statistics.
“It is a proven approach that enables discoveries to be made ten times quicker.
Once the model is built, Intellegens will work with The Richard Nixon Foundation to roll it out as a public policy tool.
Initially, it will be targeted at health and social care providers, helping them to ensure adequate resources and supplies are in place for the sick and vulnerable.
“We are confident that Alchemite will prove useful in the ongoing battle against Covid-19.
“With Innovate UK funding in place, we will create a tool that will enable governments, policy makers, care providers, and companies to better meet peak demand, whether planning the availability of critical supplies or deploying staff,” said Gareth.
Intellegens will train its Alchemite engine using online repositories and data from organisations such as the NHS, the European Centre for Disease Prevention, and the World Health Organization.
The data fed in will include:
- Real time patient case metrics including numbers of infections, recoveries, and deaths;
- Population metrics such as age distribution, housing density, and connectedness;
- Fraction of population believed to have been infected but not tested;
- Number of unreported deaths outside the hospital system;
- Statistics from limited testing facilities;
- Profiling of most at risk groups based on age, ethnicity, or other conditions; and
- Environmental, geographical and economic indicators.
The University of Cambridge spin-out has an AI toolset designed to train neural networks from sparse or noisy data that is as little as 0.05 per cent complete.
During the Covid-19 crisis, it has helped scientists, researchers and engineers conduct virtual, data-driven experiments that would otherwise have been delayed with the closure of universities, laboratories, and manufacturing plants.
Alchemite has been used in material and chemical design optimisation problems to help organisations solve data analysis bottlenecks.
Sign up to our weekly newsletter for a digest of the best stories straight to your inbox.