Data for: Improved accuracy in OCT diagnosis of rare retinal disease using few-shot learning with generative adversarial networks

Name: Data for: Improved accuracy in OCT diagnosis of rare retinal disease using few-shot learning with generative adversarial networks
Creator: TaeKeun Yoo
Published: 2020-03-27T15:17:40.056Z
Keywords: Retina, Optical Coherence Tomography

Yoo, TaeKeun

doi:10.17632/btv6yrdbmv.1

Data for: Improved accuracy in OCT diagnosis of rare retinal disease using few-shot learning with generative adversarial networks

Published: 27 March 2020| Version 1 | DOI: 10.17632/btv6yrdbmv.1

Contributor:

TaeKeun Yoo

Description

The retinal OCT images ofr rare diseases were extracted by using the Google image and Google dataset search that included English keywords including central serous chorioretinopathy, macular telangiectasia, macular hole, Stargadt’s disease, retinitis pigmentosa. These rare diseases were selected according to a previous review article about OCT diagnosis. The images possessing rare diseases were manually classified by two board-certified ophthalmologists, and ambiguous images were removed to clarify the image domains. Additional file "Segmentation_manual.zip" offers manually segmentedOCT images for pathologic lesions.

Files

Institutions

Aerospace Medical Center

Data for: Improved accuracy in OCT diagnosis of rare retinal disease using few-shot learning with generative adversarial networks

Description

Files

Institutions

Categories

Licence