Title Pilvinės aortos aneurizmos automatinės detekcijos nauda matuojant aneurizmos tūrį. Literatūros apžvalga ir modelio sukūrimas /
Translation of Title The value of automatic abdominal aortic aneurysm detection for aortic volume measurement. literature review and volume model development.
Authors Račytė, Austėja
Full Text Download
Pages 28
Abstract [eng] For almost 20 years, aortic diameter has been the main parameter used not only for diagnosing aortic aneurysms, but also for surveillance, clinical decision-making and follow-ups after interventions. Despite its widespread use, diameter measurement has some relevant limitations. Therefore, aortic aneurysm volume measurement has recently been reported as a superior and more accurate method than diameter. Although volume measurement excels in detecting aneurysm sac changes post-endovascular interventions and more accurately reflects aneurysm growth, it still not used in daily clinical practice due to technical limitations. Given that high time consumption is considered the primary drawback of volume measurement, leading to a mismatch between workload and workforce, this study aimed to compare the time consumption of manual and fully automatic aortic aneurysm volume measurement. Ten sets of abdominal and pelvic computed tomography angiography (CTA) images from 10 patients with abdominal aortic aneurysms (totaling 20 CTA image series) were utilized. Aortic and common iliac artery volumes were calculated manually in all patients using the 3mensio and OsiriX software and automatically using the EndoArt artificial intelligence algorithm. The time consumption for each measurement was recorded and then compared. Statistically significantly higher time costs were observed for manual volume calculation compared to fully automatic measurement (p<0.001, based on Wilcoxon's criterion). These findings highlight the benefits of automatic aortic aneurysm volume measurement. Implementing automatic volume measurement in clinical practice would save time, mitigate the mismatch between workload and workforce, and enable more accurate detection of small yet clinically relevant aneurysm size changes.
Dissertation Institution Vilniaus universitetas.
Type Master thesis
Language Lithuanian
Publication date 2024