University Hospital of Brussels (UZ Brussel) will be one of the first hospitals in Europe to roll out new AI programmes that aim to speed up the detection of nine acute conditions.
CT or MRI scans generate generates thousands of images per examination, providing a lot of insight but also a huge amount of information for radiologists to analyse. Reviewing the images is enormously time-consuming, the hospital's head of radiology Johan de Mey explains: "Sometimes there are 20,000 to 30,000 images per patient; they all have to be viewed by a radiologist."
"This AI technology allows radiologists to focus attention on the images where an abnormality can be seen." UZ Brussel has been working with Aidoc, a leading provider of artificial intelligence (AI) solutions for healthcare, for about four years.
The AI software can identify abnormalities, such as bleeding in the brain. This information is sent to the doctor to make a targeted diagnosis.
Providing better care
This gives a "great added value" to patients in emergency conditions said Hans Nieboer, who is in charge of Emergency Radiology at the hospital. "This way, doctors have more time for the concrete interpretation of the detected abnormalities."
"The technology helps us, especially with patients who need rapid diagnosis and intervention such as bleeding in the brain or air in the abdomen."
- AI hub to boost economic and environmental innovation opens in Brussels
- AI software ChatGPT almost smart enough to pass tough medical exam
Until recently, UZ Brussel mainly used AI to analyse CT scans for intracranial haemorrhages and pulmonary embolisms. The new system will allow for even more conditions to be detected – most of them based on CT images. The ailments that might be caught include: blockages of large blood vessels, bulges of blood vessels in the brain (aneurysms), spinal fractures with displacement of vertebrae, rib fractures, free air in the abdomen, and via RX images also pneumothorax.