No abstract

No abstract

The extraction of cardiac borders, particularly, the ones related to the left ventricle (LV), is an important goal to estimate some indices of great clinical value, such as, the thickness of the wall, ejection fraction, and regional wall motion, as the most used to assess the LV function. The accuracy of those indices depends on the correct LV boundary extraction. In this work, two LV segmentation algorithms are implemented: differencing method applied to the intensity profiles and the windows adaptive thresholds by Otsu algorithm. Results provided by the two techniques will be analysed considering factors like accuracy in the boundary extraction, effect of some artifacts like papillary muscles, intra-cavity structures, and valves, epicardial border identification, processing time. Finally, the matching between the automatic border tracing and the true anatomical border, extracted by an expert, is analysed.

Abstract. In this work we reconstruct the 3-D shape and location of the retinal vascular network from commercial spectral-domain (SD) optical coherence tomography (OCT) data. The 2-D location of retinal vascular network on the eye fundus is obtained through SVM classification of properly defined fundus images from OCT data, 1 taking advantage of the fact that on standard SD-OCT the incident light beam is absorbed by haemoglobin, therefore creating a shadow cast on OCT signal below each perfused vessel. The depth-wise location of the vessel is obtained as the beginning of the shadow. The classification of crossovers and bifurcations within the vascular network is also addressed. We illustrate the feasibility of the method in terms of vessel calibre estimation and the accuracy of bifurcations and crossovers classification.

No abstract