The paper evaluates domain generalization strategies for optic disc and cup segmentation in fundus images, highlighting issues with existing methods when applied to uncropped images, and proposes a semi-supervised learning approach based on the Noisy Student framework to improve performance across diverse datasets.
We experimentally validate whether using coarse-to-fine models instead of one-stage models is appropriate or not for segmenting the optic disc and the optic cup in color fundus images. We observed that models trained with the right amount of data can perform much better than coarse-to-find approaches.
Cuando aplicamos inteligencia artificial en medicina, solemos seguir uno de dos enfoques: o bien utilizar modelos que ya existen para resolver un problema nuevo, o bien proponer un modelo nuevo para resolver problemas que ya existen. En todos los …
Voy a presentar algunas de las líneas de trabajo actuales en inteligencia artificial aplicada a la oftalmología que llevo adelante, repasando además algunos desarrollos previos en los que participé durante mi estadía postdoctoral en Austria.
I will present some of the conclusions of the REFUGE challenge on glaucoma assessment from fundus pictures, and my personal view on what's still needed to progress towards automated screening of glaucoma from fundus pictures.