A Novel Platform for Crowd-Sourcing Retinal Image Segmentations

Jonathan Fhima1, Jan Van Eijgen2, Moti Freiman1, Ingeborg Stalmans2, Joachim A. Behar3
1Faculty of Biomedical Engineering, Technion-IIT, 2Research Group Ophthalmology, Department of Neurosciences, 3Technion-IIT


Introduction: For supervised deep learning (DL) tasks, researchers need a large annotated dataset. In medical data science, one of the major limitations to develop DL models is the lack of annotated examples in large quantity. This is most often due to the time and expertise required to annotate. We introduce, a novel platform for facilitating and crowd-sourcing image segmentations.

Methods: is composed of three components; an iPadOS client application named, a back-end server named and a python API name was developed in Swift 5.6 and is a firebase backend. allows the management of the database. can be in- stalled on as many iPadOS devices as needed so that annotators may be able to perform their segmentation simultaneously and remotely. We incorporate Apple Pencil compatibility, making the segmentation faster, more accurate, and more intuitive for the expert than any other computer-based alternative.

Results: We demonstrate the usage of for the creation of a retinal fundus dataset with reference vasculature segmentations.

Discussion and future work: We will use active learning strategies to continue enlarging our retinal fundus dataset by including a more efficient process to select the images to be annotated and distribute them to annotators.