Image Annotation Software for Artificial Intelligence Applications

Abstract

Introduction: The growing trend in artificial intelligence has recently highlighted the demand for user-friendly and effective annotation tools for researchers. Therefore, we conducted a review to assess existing annotation software that has been used in ophthalmology projects and/or available on the web. Methods: We systematically searched for AI ophthalmology studies using annotation software on PubMed on 8th July 2022 with specific criteria. Only original English articles related to ophthalmic AI were considered. From these, we identified annotation software used and conducted a subsequent Google search for additional software. Each software was evaluated based on factors like development year, accessibility, and citations of its original paper. Practicality criteria for the software included independence from external libraries, size under 100 MB, cost, and versatility in image input and output formats. Results: We identified 131 image annotation software, of which 10 met our criteria. Among the software tools utilized for image annotation in ophthalmology papers, only CVAT and ImageJ were freely accessible. This paper provides a concise overview of the 10 image annotation software. Conclusions We systematically analyzed annotation software for fundus image annotation, highlighting 10 primary tools with varied functionalities. However, this study is limited to AI-related software, underscoring the need for continual updates due to the evolving nature of image annotation tools.

Ähnliche Arbeiten