Covid-19: NUIG project to help rapidly identify seriously-ill patients
Scientists to use national supercomputer and artificial intelligence to develop system
The system will quickly process large volumes of chest images so deteriorating Covid-19 patients can be identified when admitted to hospital. Photograph: iStock
A team of scientists based at NUI Galway is set to use the national supercomputer, known as Kay, to help develop a system to quickly process large volumes of chest images so deteriorating Covid-19 patients can be identified when admitted to hospital.
They will be using artificial intelligence (AI) techniques to evaluate computerised tomography (CT) scans of the chest – as with many seriously-ill people who may have Covid-19, there is often insufficient time to do a PCR test to confirm diagnosis.
With the ongoing pandemic, the Irish Centre for High-End Computing (ICHEC) expedited approval for the project using the high-performance computer, Kay, which is based at Waterford Institute of Technology.
The project is being undertaken by Dr Aaron Golden of NUIG school of mathematics, Dr Christoph Kleefeld of NUIG school of physics and Dr Declan Sheppard, clinical director of radiology at Galway University Hospitals. It is funded by the Health Research Board under its rapid Covid-19 funding programme.
“Everyone is used to the idea of nasal and throat swabs being used as the standard test for Covid, but they’re not 100 per cent accurate. In fact, they can miss genuine cases and there can also be significant delays in getting the result back,” said Dr Golden.
With patients who have underlying conditions or someone showing severe signs of distress, being able to quickly determine the presence and stage of pulmonary disease is critical.
“A quick and non-invasive means of determining this is via a CT scan. In less than an hour a radiologist can be in a position to see evidence of lesions on the lungs that would be indicative of Covid-19,” he added.
It can be difficult to tell whether a lesion is caused by community-acquired pneumonia, other pulmonary disorders or other common lung conditions. “However, an imaging system that incorporates AI can be trained to distinguish between them and deployed to assist the radiology team,” Dr Golden said.
A phenomenal global effort is ongoing to put in place “a state-of-the-art AI machine” using software stacks that can be tuned by research groups to classify patient CT scans into likely-Covid or non-Covid groups, he said.
“Tuning makes it sound like a minor thing, but . . . we’re talking about training AI systems over and over again with thousands of different CT scans.”
They will also use deep learning algorithms to standardise and train thousands of CT scans to eliminate “biases and artefacts that are a consequence of scans originating from different imaging centres taken using different settings”. This is a big data problem requiring a phenomenal amount of computation and where Kay’s computational power makes all the difference.
Dr Golden has worked in the United States using supercomputing resources there. “I have to say that the way ICHEC has optimised user access to Kay is as good as anything I previously used in the US, if not better. It is a fantastic asset for the health research community here in Ireland. ”