Supervised Segmentation of Un-Annotated Retinal Fundus Images by Synthesis
Open Access
- 24 July 2018
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Imaging
- Vol. 38 (1) , 46-56
Abstract
We focus on the practical challenge of segmenting new retinal fundus images that are dissimilar to existing well-annotated datasets. It is addressed in this paper by a supervised learning pipeline, with its core being the construction of a synthetic fundus image dataset using the proposed R-sGAN technique. The resulting synthetic images are realistic-looking in terms of the query images, while maintaining the annotated vessel structures from the existing dataset. This helps to bridge the mismatch between the query images and the existing well-annotated dataset. As a consequence, any known supervised fundus segmentation technique can be directly utilized on the query images, after training on this synthetic dataset. Extensive experiments on different fundus image datasets demonstrate the competitiveness of the proposed approach in dealing with a diverse range of mismatch settings.Keywords
Funding Information
- A*STAR JCO Grants
- Chinese Government Scholarship
This publication has 24 references indexed in Scilit:
- A Cross-Modality Learning Approach for Vessel Segmentation in Retinal ImagesIEEE Transactions on Medical Imaging, 2015
- Biologically-Inspired Supervised Vasculature Segmentation in SLO Retinal Fundus Images2015
- An effective retinal blood vessel segmentation method using multi-scale line detectionPattern Recognition, 2013
- Blood vessel segmentation methodologies in retinal images – A surveyComputer Methods and Programs in Biomedicine, 2012
- Fast Retinal Vessel Detection and Measurement Using Wavelets and Edge Location RefinementPLOS ONE, 2012
- A review of vessel extraction techniques and algorithmsACM Computing Surveys, 2004
- Ridge-Based Vessel Segmentation in Color Images of the RetinaIEEE Transactions on Medical Imaging, 2004
- Segmentation of vessel-like patterns using mathematical morphology and curvature evaluationIEEE Transactions on Image Processing, 2001
- Locating blood vessels in retinal images by piecewise threshold probing of a matched filter responseIEEE Transactions on Medical Imaging, 2000
- Long Short-Term MemoryNeural Computation, 1997