ENGINEERINGNET.BE – A CT or MRI scan generates thousands of images per examination. This means that radiologists have to analyze an enormous amount of radiological images per patient. Today they have less than 1 second per image to view this amount of images of their patients.
UZ Brussel is working with Aidoc, a leading provider of AI solutions for healthcare, to improve this process.
With the help of AI-based software, images are forwarded and analyzed ‘pseudonymised’. The software can identify abnormalities, for example a hemorrhage in the brain. The doctor receives this information and can thus analyze the images in a targeted manner and make a diagnosis.
Until recently, UZ Brussel mainly relied on AI for the analysis of CT scans for intracranial bleeding and pulmonary emboli.
Now even more conditions can be targeted, mostly based on CT images: blockages of large blood vessels, bulges of the blood vessels in the brain, spinal fractures with displacement of vertebrae, rib fractures, free air in the abdomen, and through RX images also a collapsed lung.
This means that half the number of patients can now be helped more quickly. This makes UZ Brussel the first hospital to use all Aidoc AI solutions for the triage of acute conditions.