COVID-19 training tool for lung CT scans could help diagnosis

by | Coronavirus - COVID-19

31 March 2020

Sydney scientists have developed a free online program to train doctors and radiologists to identify signs of COVID-19 in lung CT scans.

CT scans are used to view the extent of COVID-19 within the lungs. As demand ramps up, there may be a shortage of healthcare workers experienced in interpreting lung CT scans. The new platform, named CovED, can quickly upskill staff unfamiliar with lung radiology to prepare standardised reports for experts to review.

CovED has been developed by DetectED-X, a Sydney start-up which specialises in radiation and imaging. DetectED-X has already developed a breast cancer tool and in 2019 were commissioned by the Australian Government to deliver a solution for diagnosing dust disease.

Medical radiation scientist, educator and CEO of DetectED-X, Professor Patrick Brennan of the University of Sydney, said early and better diagnosis would help relieve overburdened healthcare systems and save lives.

“The number of patients that are suffering from this life-threatening illness is fast outpacing the number of skilled staff required to accurately diagnose the required lung CT scans,” Professor Brennan said.

The platform has been made available for free worldwide and uses algorithms to improve the radiologist’s skills and ability to identify COVID-19 in lung CT scans. The platform uses CT scans from Australian patients with COVID-19.

Some of the main signs of COVID-19 on lung CT scans are:

  • ‘ground-glass opacity’ – a hazy appearance (instead of the normal black) that doesn’t obscure the underlying lung structures. It indicates a partial filling of the lung’s air spaces with fluid
  • ‘crazy paving’ – where there are extensive ground-glass opacities and a pattern that looks like crazy paving
  • consolidation – where air spaces are filled with liquid (e.g. mucus) and appear white.

These signs are not limited to COVID-19.