Classification of SD-OCT Volumes with LBP: Application to DME Detection
- Resource Type
- Authors
- Guillaume Lemaitre; Mojdeh Rastgoo; Joan Massich; Shrinivasan Sankar; Fabrice Meriaudeau; Désiré Sidibé
- Source
- Ophthalmic Medical Image Analysis Workshop (OMIA), Medical Image Computing and Computer Assisted Interventions (MICCAI) 2015
Ophthalmic Medical Image Analysis Workshop (OMIA), Medical Image Computing and Computer Assisted Interventions (MICCAI) 2015, Oct 2015, Munich, Germany. 2015
Ophthalmic Medical Image Analysis Workshop (OMIA), Medical Image Computing and Computer Assisted Interventions (MICCAI) 2015, Oct 2015, Munich, Germany
- Subject
- genetic structures
Local binary patterns
Computer science
Diabetic macular edema
Spectral domain
02 engineering and technology
[ SPI.SIGNAL ] Engineering Sciences [physics]/Signal and Image processing
030218 nuclear medicine & medical imaging
03 medical and health sciences
0302 clinical medicine
Optical coherence tomography
Discriminative model
LBP
0202 electrical engineering, electronic engineering, information engineering
medicine
DME
Computer vision
[SPI.SIGNAL] Engineering Sciences [physics]/Signal and Image processing
medicine.diagnostic_test
business.industry
eye diseases
Diabetic Macular Edema
OCT
020201 artificial intelligence & image processing
Artificial intelligence
sense organs
Optical Coherence Tomography
business
[SPI.SIGNAL]Engineering Sciences [physics]/Signal and Image processing
- Language
- English
International audience; This paper addresses the problem of automatic classification of Spectral Domain OCT (SD-OCT) data for automatic identification of patients with Diabetic Macular Edema (DME) versus normal subjects. Our method is based on Local Binary Patterns (LBP) features to describe the texture of Optical Coherence Tomography (OCT) images and we compare different LBP features extraction approaches to compute a single signature for the whole OCT volume. Experimental results with two datasets of respectively 32 and 30 OCT volumes show that regardless of using low or high level representations, features derived from LBP texture have highly discriminative power. Moreover, the experiments show that the proposed method achieves better classification performances than other recent published works.