Scientists from Gdańsk University of Technology, the Medical University of Gdańsk and the University Clinical Center in Gdańsk are building software designed to estimate both the likelihood that a patient has an intracranial aneurysm and the risk that it could rupture.
The system is meant to work by analysing routine laboratory results, a patient’s medical history, and text from clinical documentation, according to Gdańsk University of Technology spokesman Patryk Rosiński.
"This is a breakthrough approach," Rosiński said, arguing that until now such assessments were widely seen as possible only through imaging tests such as computed tomography scans or angiography.
The project, titled "AI-Powered Medical Software for Predicting the Likelihood of Intracranial Aneurysm," has been selected for the international I3HIES accelerator program, which supports promising healthcare innovations on their path toward investment and commercial use.
The project is planned in three stages, the researchers said. The first focuses on predicting aneurysm rupture risk. The second aims to identify whether a patient belongs to a higher-risk group. The third is expected to deliver a clinical calculator and an app for physicians to support treatment decisions for patients considered at risk.
The team trained its early models using data from more than 60,000 patients treated at the University Clinical Center in Gdańsk between 2006 and 2024.
Rosiński said the first results were encouraging, reporting effectiveness of more than 77 percent and sensitivity of about 80 percent, figures that suggest potential clinical usefulness if confirmed in further work.
Project leader Patryk Jasik of the Gdańsk University of Technology said the team used modern predictive models and multi-criteria validation to avoid inflated performance, adding that some analyses exceeded 90-percent accuracy.
He also said the dataset was prepared with safeguards to prevent information leakage during modelling.
Dr. Justyna Fercho, who leads the team at the Medical University of Gdańsk and the University Clinical Center, said doctors value approaches that can be interpreted and applied in practice.
She said understanding what drives a model’s risk assessment can help clinicians diagnose patients more effectively.
The researchers pointed to the scale of the problem, with as many as one in 50 adults may have an unruptured aneurysm. This would mean hundreds of thousands of people in Poland alone could be affected.
Treatment can involve surgical clipping, which requires opening the skull, or less invasive endovascular procedures such as coiling, sometimes combined with stents or flow-diverting devices for more complex cases.
The team said costs rise sharply after a rupture, when patients often need lengthy rehabilitation and may be unable to return to work.
Jasik said earlier detection through risk assessment could save lives and reduce strain on the healthcare system.
(rt/gs)
Source: PAP, wnp.pl, naukawpolsce.pl