TY - GEN
T1 - Optic Disk Detection in Fundus Images of Retinopathy of Prematurity
AU - Intriago-Pazmiño, Monserrate
AU - Ibarra-Fiallo, Julio
AU - Pérez-Hernández, María
AU - Guzmán-Castillo, Adán
AU - Torres-Constante, Eddy
N1 - Publisher Copyright:
© 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2022/10/14
Y1 - 2022/10/14
N2 - Children born prematurely may suffer from an eye disease called retinopathy of prematurity (ROP). To estimate the severity of this disease, physicians need to type, among other things, the extent of the disease on images that are often of poor quality. The extent of ROP is measured from the optical disc. Therefore, it is essential to have an automatic method that locates and segments the optical disc. In order to contribute to a computational method that detects the optic disc in children’s pathological images, a fast-processing method is presented in this work. This method creates a template-based local binary pattern histogram. Next, the method proposes recognizing candidate windows as an optical disk from regional maxima. Then, the template is used to choose the correct optic disk. This method used thirty images from the ROPFI dataset that contains infant pathological images. The optic disk has been manually labeled. The test identifying the optic disk achieved a sensitivity of 0.95.
AB - Children born prematurely may suffer from an eye disease called retinopathy of prematurity (ROP). To estimate the severity of this disease, physicians need to type, among other things, the extent of the disease on images that are often of poor quality. The extent of ROP is measured from the optical disc. Therefore, it is essential to have an automatic method that locates and segments the optical disc. In order to contribute to a computational method that detects the optic disc in children’s pathological images, a fast-processing method is presented in this work. This method creates a template-based local binary pattern histogram. Next, the method proposes recognizing candidate windows as an optical disk from regional maxima. Then, the template is used to choose the correct optic disk. This method used thirty images from the ROPFI dataset that contains infant pathological images. The optic disk has been manually labeled. The test identifying the optic disk achieved a sensitivity of 0.95.
KW - Medical image processing
KW - Optic disk detection
KW - Retinopathy of prematurity
UR - https://www.scopus.com/pages/publications/85142074498
U2 - 10.1007/978-3-031-18344-7_25
DO - 10.1007/978-3-031-18344-7_25
M3 - Contribución a la conferencia
AN - SCOPUS:85142074498
SN - 9783031183430
T3 - Lecture Notes in Networks and Systems
SP - 370
EP - 380
BT - Proceedings of the Future Technologies Conference, FTC 2022, Volume 3
A2 - Arai, Kohei
PB - Springer Science and Business Media Deutschland GmbH
T2 - 7th Future Technologies Conference, FTC 2022
Y2 - 20 October 2022 through 21 October 2022
ER -