Abdominal aortic aneurysm segmentation
We have developed software for abdominal aortic aneurysm segmentation from CT images. The software includes new deep learning-based methods for the automatic extraction of different types of abdominal aortic aneurysms, has a high segmentation accuracy and its computation takes only a few seconds.
Centreline tracking of the single coronary artery from X-ray angiograms
We have developed software for semi-automatic centreline extraction and tracking of the single coronary artery. Our approach is based on ridge edge detection and thresholding, after which the centreline of a vessel tree is extracted using skeletonization. After that, we manually select two points: at the start and the end of the single coronary artery. Finding the shortest path between selected points is used for single coronary centreline extraction. Finally, the template matching of selected points allows centreline tracking through the whole angiogram sequence.
