LUMC calculates lung damages COVID-19 patients with AI technology

Healthcare professionals at Leiden University Medical Centre (LUMC) can use artificial intelligence (AI) to quickly and easily calculate whether a corona patient has suffered serious lung damage. They do this by running a CT scan through the AI software CAD4COVID-CT. The software was developed by the company Thirona and is used in the LUMC via Philips Intellispace AI Workflow Suite.

Corona patients who are admitted to the hospital usually get a CT scan taken to determine the extent of infection of the lungs. The assessment of these scans is now automated with the AI ​​software CAD4COVID-CT. The software determines the severity of COVID-19 disease based on the percentage of affected lung tissue. This saves valuable time for doctors and radiologists.

The application of AI is part of the Clinical Artificial Intelligence Implementation and Research Lab (CAIRELab). This is a knowledge and expertise center in the LUMC in which AI methods are brought together. As a partner in CAIRELab, the Radiology department is one of the most advanced users of AI. With the data from the quantitative image analysis, healthcare professionals can quickly make an accurate diagnosis of the severity of the disease. In addition, they can also better predict disease progression.

“The precision of CAD4COVID is helpful in diagnosing the severity and progression of the COVID-19 disease,” says Hildo Lamb, radiologist at LUMC. “A radiologist’s ‘human intelligence’ is well equipped to recognise patterns, but extremely precise quantification and calculations of asymmetric areas are more difficult. Having Artificial Intelligence support these measurements saves time, and boosts precision. Hence, delivering a faster and more accurate diagnosis of the COVID-19 disease, which enables precision diagnosis and monitoring of treatment of COVID-19 patients.”

Radiologist is always watching

After each calculation, a PDF report will appear in the system. These are all viewed and assessed by a radiologist at the LUMC. "The report contains a quantification of how much lung tissue has been damaged in the different lung lobes," says Hildo Lamb. "In addition, the system creates an ‘overlay image’, so that the radiologist can see on what the algorithm based the choice."

Big step for AI in healthcare

CAD4COVID-CT was developed by Thirona in response to the corona pandemic. “We wanted to make a contribution,” says Mark van Grinsven, head of product development at Thirona. “We are happy to partner with Philips to bring AI to the forefront of healthcare and provide the users at LUMC a fully integrated workflow. We believe this is a great first step towards automatic analysis of medical images with AI.”

The LUMC is the first hospital in the world to introduce IntelliSpace AI workflow suite. With the IntelliSpace AI workflow suite, healthcare providers can integrate AI applications into their work process.