Mar 2020 – While the latest development in AI has been proven to be relatively accurate detecting retinal diseases, there has been little insight on what impacts clinical deployment in real-world settings. In a recent publication in NPJ Digital Medicine, EyRIS founders shared their findings from a systematic review of literature on possible factors, both image-related and technical affecting the performance of deep learning algorithms for retinal image analysis.
The review covered 202 publications.
Some of the technical parameters include choice of convolutional neural network architecture and computational framework while image-related considerations included type of retinal camera, ethnic group, training data set size, image compression, etc.
This article was originally published on NPJ Digital Medicine on 23 Mar 2020 and can be read at: https://www.nature.com/articles/s41746-020-0247-1 .
About NPJ Digital Medicine
NPJ Digital Medicine is part of Nature Research which serves the research community by publishing its most significant discoveries—findings that advance knowledge and address some of the greatest challenges that we face as a society today. NPJ Digital Medicine is an international, peer-reviewed journal dedicated to publishing the highest quality research relevant to all aspects of digital medicine and health.
Please direct all queries related to this article to Marketing at firstname.lastname@example.org