automated drusen detection and quantificatioon system
TRANSCRIPT
By : Gehad Hassan
AUTOMATED DRUSEN DETECTION
AND QUANTIFICATION SYSTEM
Fayoum University
Dept. of Computer Science, Faculty of Computers and information
Member of the Scientific Research Group in Egypt
Agenda
Introduction.
Problem Definition
Main Objectives
Drusen Imaging Ways
Retinal Layers of A cross-sectional SD-OCT Image (B-scan).
Problems in Consideration
The Proposed Recommender System
Introduction
Drusen are tiny yellow or white deposits in a layer of the
retina called Bruchs membrane.
They are the most common early sign of dry age-related
macular degeneration(AMD).
Introduction Cont..
Drusen are made up of lipids, a type of fatty protein. They may be the result of a failure of the eye to dispose of waste products that are produced when the photoreceptors of the eye drop off older parts of the cell.
Introduction Cont..
There are several types of drusen with different levels of risk.
Drusen can be small, hard and scattered far apart from each other.
Some drusen can become larger, softer and closer together.
Introduction Cont..
An estimated 8 million persons at least 55 years old in the United States have monocular or binocular intermediate AMD or monocular advanced AMD (In, 2003). One of its clinical characteristics and, in most cases, the first clinical finding is the presence of drusen.
it is considered the injury of Drusen of the biggest causes of blindness in the whole world, and the incidence of drusen constitutes 75% of the causes of blindness, while other diseases (blue water - cloud the cornea - the network separation - bleeding eye, etc. ..) make up 25% of the causes of blindness .
Drusen Imaging Ways
Drusen Imaging Ways
CFP(color fundus photographs )
OCT(Optical coherence tomography)
Problem Definition
Detection of drusen is very important issue in
medical field to avoid some problems that affects
central vision or it can be used for the assessment of
change in disease status - response to treatment or
progression . So the detection of drusen part and
the calculation of its characteristics may help
doctors to determine diseases status.
Main Objectives
Developing an effective automated drusen
segmentation System.
Extracting qualitative features from drusen which
are useful for evaluating the progress of these
lesions.
Problems in consideration
This pervious methods have some limitation
RPE and IS OS have the same reflectivity Limitations in small
drusen segmentation
Retinal layers of a cross-sectional SD-
OCT image (B-scan)
RNFL: retinal nerve fiber layer
IS/OS: photoreceptor inner/outer
segments.
RPE: retinal pigment epithelium. The
location of a druse is indicated with a
yellow arrow.
The RNFL complex (is indicated by the
orange dotted region)
Retinal layers of a cross-sectional SD-
OCT image (B-scan) Cont..
more effect on
vision
More Risk of AMD
Bigger area and
size of Drusen
Input image
OCT image
Image
denoising
1
RPE
segmentation
2
Drusen
segmentation
4
Drusen
projection
5 Elimination of
false positive
drusen
6
Drusen
refinemnton
projection
image
7
Drusen
smoothing
8
RNFL Layer
removal
3
The Proposed
Recommender
System
Output
Drusen
category status
of disease
Features Extraction