Article Text

Download PDFPDF
Original research
Artificial intelligence aneurysm measurement tool finds growth in all aneurysms that ruptured during conservative management


Background Cerebral aneurysm rupture is associated with high rates of morbidity and mortality. Detecting aneurysms at high risk of rupture is critical in management decision making. Rupture risk has traditionally been associated with size—measured as a maximum dimension. However, aneurysms are morphologically dynamic, a characteristic ignored by large prospective aneurysm risk studies. Manual measurement is challenging and fraught with error. We used an artificial intelligence (AI) measurement tool to study aneurysms that ruptured during conservative management to detect changes in size not appreciated by manual linear measurement.

Methods A single practice database with >5000 aneurysms was queried. Patients followed conservatively for an unruptured aneurysm were identified using appropriate diagnosis codes. This cohort was screened for subsequent rupture using procedure codes. Only patients with two vascular imaging studies before rupture were included.

Results Five patients met the criteria. All patients had aneurysm enlargement, two of which were not detected from manual linear measurements, including adjudication and analysis, during a multidisciplinary neurovascular conference in a high volume practice. Maximum dimension increased at a minimum of 1.8% (range 1.8–63.3%) from the first scan to the last, and aneurysm volume increased at a minimum of 5.9% (5.9–385.5%), highlighting the importance of volumetric measurement.

Conclusions AI-enabled volumetric measurements are more sensitive to changes in size and detected enlargement in all aneurysms that ruptured during conservative management. This finding has major implications for clinical practice and methods used for interval aneurysm measurement in patients being conservatively followed.

  • Aneurysm
  • Intervention

Data availability statement

Data are available upon reasonable request. Data that support this study is available from the corresponding author upon reasonable request.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles