Nima Hatami, Ph.D.

PANDORE: Protection against new kind of reverse engineering

I) The proposed time-series classification system

Fusion of Local Colour-Texture Image Descriptors for Melanoma Diagnosis

Example of Melanoma and non-Melanoma (benign) samples.

Example of challenging images for segmentation.

Bag of Visual Word (BoW) classification.

Structured Analysis of the Retina

The STARE (STructured Analysis of the REtina) project was first funded by the U.S. National Institutes of Health (NIH) at the University of California San Diego (UCSD). The main idea is to take advantage of machine's power to provide information needed in ophthalmology for automatic diagnosis and treatment purposes. Images and clinical data were provided by the UCSD based Shiley Eye Center, and by the Veterans Administration Medical Center in San Diego. The problems I am currently working on:

i) Automatic Identification of Retinal Arteries and Veins in Fundus Images (left: hand-labeled image by the gold standard; right: results of our proposed automatic system)

ii) Automatic retinal layer segmentation of OCT (results of our proposed segmentation algorithm)

iii) Retinal blood vessel tortuosity measurement