DocumentCode :
1707569
Title :
A digital method of drusen quantification based on the geometry of fundus reflectance
Author :
Smith, R. Theodore ; Chan, Jackie K. ; Nagasaki, Takayulu ; Sparrow, Janet R. ; Barbazetto, Irene ; Klaver, Caroline C W
Author_Institution :
Dept. of Ophthalmology, Columbia Univ., New York, NY, USA
fYear :
2003
Firstpage :
91
Lastpage :
92
Abstract :
An interactive semi-automated procedure was developed to level the macular background reflectance for the purpose of morphometric analysis of drusen. In this study, 10 color fundus photographs of patients with age related macular degeneration (AMD) and drusen were used. After digitizing the photographs, the underlying background pattern in the green channel was leveled by a stepwise algorithm based on a geometric model of reflectance in the normal fovea and parafovea. Segmentation of drusen and measurements were made in the central and middle subfields (1000 μm diameter circle and 3000 μm diameter annulus). Two graders independently applied the digital method to the ten cases. The mean digital measurements were compared to stereo fundus gradings by an expert grader. The absolute mean differences between the digital and stereo gradings were 2.4 +/- 2.8% in the middle subfield and 2.6 +/- 3.1% in the central subfield. Two digital graders using the software achieved comparable drusen measurements to stereo gradings. This technique for macular image analysis has the potential of being adopted in large clinical trials.
Keywords :
biomedical optical imaging; colour photography; eye; geriatrics; image colour analysis; image segmentation; mathematical morphology; medical image processing; optical information processing; reflectivity; 1000 micron; 3000 micron; absolute mean differences; age related macular degeneration patients; background pattern; blindness; central subfields; color fundus photographs; digital method; drusen quantification; expert grader; fundus reflectance geometry; geometric model; green channel; interactive semi-automated procedure; large clinical trials; macular background reflectance; macular image analysis; mean digital measurements; middle subfields; morphometric analysis; normal fovea; parafovea; segmentation; stepwise algorithm; stereo fundus gradings; stereo gradings; Brightness; Geometry; Image color analysis; Image segmentation; Pathology; Pixel; Reflectivity; Software measurement; Solid modeling; Testing;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Bioengineering Conference, 2003 IEEE 29th Annual, Proceedings of
Print_ISBN :
0-7803-7767-2
Type :
conf
DOI :
10.1109/NEBC.2003.1216007
Filename :
1216007
Link To Document :
بازگشت