Research Article
Training U-Net with Proportional Image Division for Retinal Structure Segmentation
@INPROCEEDINGS{10.1007/978-3-031-60665-6_9, author={Pedro Victor de Abreu Fonseca and Alexandre Carvalho Ara\^{u}jo and Jo\"{a}o Dallyson S. de Almeida and Geraldo Braz J\^{u}nior and Arist\^{o}fanes Correa Silva and Rodrigo de Melo Souza Veras}, title={Training U-Net with Proportional Image Division for Retinal Structure Segmentation}, proceedings={Wireless Mobile Communication and Healthcare. 12th EAI International Conference, MobiHealth 2023, Vila Real, Portugal, November 29-30, 2023 Proceedings}, proceedings_a={MOBIHEALTH}, year={2024}, month={6}, keywords={Segmentation Retinal fundus Proportional Image Division U-Net}, doi={10.1007/978-3-031-60665-6_9} }- Pedro Victor de Abreu Fonseca
Alexandre Carvalho Araújo
João Dallyson S. de Almeida
Geraldo Braz Júnior
Aristófanes Correa Silva
Rodrigo de Melo Souza Veras
Year: 2024
Training U-Net with Proportional Image Division for Retinal Structure Segmentation
MOBIHEALTH
Springer
DOI: 10.1007/978-3-031-60665-6_9
Abstract
Cup and optic disc segmentation has become one of the main objects of study in the field of creating and improving machine learning-oriented models due to the importance of vision for human beings and the ability to assist physicians in diagnosing ocular problems. Within this context, this study presents a new method based on the proportional division of images concerning features extracted from the sample set. These samples go through a pre-processing step involving image resizing before going to deep feature extraction and K-means clustering, thus dividing the set for validation and training. Soon after, the amount of samples is increased through data augmentation before going on to the U-Net training. The proposed method has been evaluated on the public RIM-ONE and DRISHTI-GS datasets, and presented promising results in the segmentation of both structures, with emphasis on obtaining the value of 92.2% of Dice for the segmentation of the optic cup in the DRISHTI-GS test dataset and 95.9% of Dice for the optic disc in the RIM-ONE.
