Vanya V. Valindria

Learn More
In this study, we developed a novel method for a fullyautomatic segmentation and quantification of myocardialinfarction from delay enhancement - magnetic resonanceimaging. Hyper-enhanced region corresponding to the infarctcore is separated from normal tissue via a Gaussian mixturemodel. Then, peri-infarct area is determined by spatial-weightedfuzzy(More)
Imaging of retinal blood vessels performed by fundus fluorescence angiography (FFA) requires the use of injected fluorescence agent to enhance blood contrast during image acquisition. This procedure leads to several risks to the patient, including nausea and loss of consciousness. Another available approach which applies no contrast agent is the fundus(More)
When integrating computational tools such as automatic segmentation into clinical practice, it is of utmost importance to be able to assess the level of accuracy on new data, and in particular, to detect when an automatic method fails. However, this is difficult to achieve due to absence of ground truth. Segmentation accuracy on clinical data might be(More)
  • 1