TY - GEN
T1 - Hard Exudate Detection Using Local Texture Analysis and Gaussian Processes
AU - Colomer, Adrián
AU - Ruiz, Pablo
AU - Naranjo, Valery
AU - Molina, Rafael
AU - Katsaggelos, Aggelos K.
N1 - Funding Information:
This work has been supported in part by the Ministerio de Economía y Competi-tividad under contracts DPI2016-77869-C2-{1,2}-R, and the Department of Energy grant DE-NA0002520. The work of Adrián Colomer has been supported by the Spanish FPI Grant BES-2014-067889. We gratefully acknowledge the support of NVIDIA Corporation with the donation of the Titan Xp GPU used for this research.
PY - 2018
Y1 - 2018
N2 - Exudates are the most noticeable sign in the first stage of diabetic retinopathy. This disease causes about five percent of world blindness. Making use of retinal fundus images, exudates can be detected, which helps the early diagnosis of the pathology. In this work, a novel method for automatic hard exudate detection is presented. After an exhaustive pre-processing step, Local Binary Patterns Variance (LBPV) histograms are used to locally extract texture information. We then use Gaussian Processes to distinguish between healthy and pathological retinal patches. The proposed methodology is validated using the E-OPHTA exudates database. The experimental results demonstrate that Gaussian Process classifiers outperform the current state of the art classifiers for this problem.
AB - Exudates are the most noticeable sign in the first stage of diabetic retinopathy. This disease causes about five percent of world blindness. Making use of retinal fundus images, exudates can be detected, which helps the early diagnosis of the pathology. In this work, a novel method for automatic hard exudate detection is presented. After an exhaustive pre-processing step, Local Binary Patterns Variance (LBPV) histograms are used to locally extract texture information. We then use Gaussian Processes to distinguish between healthy and pathological retinal patches. The proposed methodology is validated using the E-OPHTA exudates database. The experimental results demonstrate that Gaussian Process classifiers outperform the current state of the art classifiers for this problem.
KW - Bayesian modeling
KW - Gaussian Processes
KW - Hard exudate
KW - Local Binary Patterns
KW - Variational inference
UR - http://www.scopus.com/inward/record.url?scp=85049441499&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85049441499&partnerID=8YFLogxK
U2 - 10.1007/978-3-319-93000-8_73
DO - 10.1007/978-3-319-93000-8_73
M3 - Conference contribution
AN - SCOPUS:85049441499
SN - 9783319929996
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 639
EP - 649
BT - Image Analysis and Recognition - 15th International Conference, ICIAR 2018, Proceedings
A2 - ter Haar Romeny, Bart
A2 - Karray, Fakhri
A2 - Campilho, Aurelio
PB - Springer Verlag
T2 - 15th International Conference on Image Analysis and Recognition, ICIAR 2018
Y2 - 27 June 2018 through 29 June 2018
ER -