Machine learning techniques for diabetic macular edema (DME) classification on SD-OCT images

Autor: Désiré Sidibé, Khaled Alsaih, Mojdeh Rastgoo, Fabrice Meriaudeau, Guillaume Lemaitre, Joan Massich
Přispěvatelé: Laboratoire d'Electronique, d'Informatique et d'Image [EA 7508] (Le2i), Université de Technologie de Belfort-Montbeliard (UTBM)-Université de Bourgogne (UB)-École Nationale Supérieure d'Arts et Métiers (ENSAM), Arts et Métiers Sciences et Technologies, HESAM Université (HESAM)-HESAM Université (HESAM)-Arts et Métiers Sciences et Technologies, HESAM Université (HESAM)-HESAM Université (HESAM)-AgroSup Dijon - Institut National Supérieur des Sciences Agronomiques, de l'Alimentation et de l'Environnement-Centre National de la Recherche Scientifique (CNRS), Centre for Intelligent Signal and Imaging Research [Petronas] (CISIR), Universiti Teknologi PETRONAS (UTP), Laboratoire d'Electronique, d'Informatique et d'Image UMR CNRS 6306 ( Le2i ), Université de Technologie de Belfort-Montbeliard ( UTBM ) -Centre National de la Recherche Scientifique ( CNRS ) -École Nationale Supérieure d'Arts et Métiers ( ENSAM ) -Université de Bourgogne ( UB ) -AgroSup Dijon - Institut National Supérieur des Sciences Agronomiques, de l'Alimentation et de l'Environnement, Centre for Intelligent Signal and Imaging Research (Universiti Teknologi Petronas) ( CISIR )
Jazyk: angličtina
Rok vydání: 2017
Předmět:
SD-OCT
lcsh:Medical technology
genetic structures
Local binary patterns
Computer science
BoW
Biomedical Engineering
02 engineering and technology
Signal-To-Noise Ratio
Macular Edema
Biomaterials
Machine Learning
03 medical and health sciences
0302 clinical medicine
Optical coherence tomography
LBP
0202 electrical engineering
electronic engineering
information engineering
medicine
Image Processing
Computer-Assisted
Humans
Radiology
Nuclear Medicine and imaging
Computer vision
[ SDV.IB ] Life Sciences [q-bio]/Bioengineering
Feature detection (computer vision)
Diabetic Retinopathy
Radiological and Ultrasound Technology
medicine.diagnostic_test
business.industry
General Medicine
Classification
DME detection
eye diseases
Support vector machine
Histogram of oriented gradients
lcsh:R855-855.5
HoG
Bag-of-words model
Feature (computer vision)
Principal component analysis
Reseach
030221 ophthalmology & optometry
020201 artificial intelligence & image processing
[SDV.IB]Life Sciences [q-bio]/Bioengineering
Artificial intelligence
sense organs
SD-OCT Classification
business
Tomography
Optical Coherence
Zdroj: BioMedical Engineering
BioMedical Engineering OnLine, BioMed Central, 2017, ⟨10.1186/s12938-017-0352-9⟩
BioMedical Engineering OnLine, Vol 16, Iss 1, Pp 1-12 (2017)
BioMedical Engineering OnLine, BioMed Central, 2017, 〈https://biomedical-engineering-online.biomedcentral.com/articles/10.1186/s12938-017-0352-9〉. 〈10.1186/s12938-017-0352-9 〉
ISSN: 1475-925X
DOI: 10.1186/s12938-017-0352-9⟩
Popis: Background Spectral domain optical coherence tomography (OCT) (SD-OCT) is most widely imaging equipment used in ophthalmology to detect diabetic macular edema (DME). Indeed, it offers an accurate visualization of the morphology of the retina as well as the retina layers. Methods The dataset used in this study has been acquired by the Singapore Eye Research Institute (SERI), using CIRRUS TM (Carl Zeiss Meditec, Inc., Dublin, CA, USA) SD-OCT device. The dataset consists of 32 OCT volumes (16 DME and 16 normal cases). Each volume contains 128 B-scans with resolution of 1024 px × 512 px, resulting in more than 3800 images being processed. All SD-OCT volumes are read and assessed by trained graders and identified as normal or DME cases based on evaluation of retinal thickening, hard exudates, intraretinal cystoid space formation, and subretinal fluid. Within the DME sub-set, a large number of lesions has been selected to create a rather complete and diverse DME dataset. This paper presents an automatic classification framework for SD-OCT volumes in order to identify DME versus normal volumes. In this regard, a generic pipeline including pre-processing, feature detection, feature representation, and classification was investigated. More precisely, extraction of histogram of oriented gradients and local binary pattern (LBP) features within a multiresolution approach is used as well as principal component analysis (PCA) and bag of words (BoW) representations. Results and conclusion Besides comparing individual and combined features, different representation approaches and different classifiers are evaluated. The best results are obtained for LBP\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$_{16 {\text{-}} \mathrm{ri}}$$\end{document}16-ri vectors while represented and classified using PCA and a linear-support vector machine (SVM), leading to a sensitivity(SE) and specificity (SP) of 87.5 and 87.5%, respectively.
Databáze: OpenAIRE
Externí odkaz: