A new feature for automatic aneurysm detection

We propose a new feature that can be used to automatically detect cerebral aneurysms in angiographic images. It combines both low-level and high-level features to a feature indicating aneurysms. The feature is used in a system for aneurysm detection in two types of magnetic resonance angiography (MRA) images and computed tomography angiography (CTA) images. The method was tested on 66 angiographic data sets containing aneurysm and non-aneurysm cases. We show that the newly introduced incorporation of the location based feature improves the detection quality. We achieve a sensitivity higher than 93% for all modalities with an average false positive rate varying from 8.8 to 20.9 per data set, depending on the modality.