Contrastive learning-based pretraining improves representation and transferability of diabetic retinopathy classification models

no code implementations • 24 Aug 2022 • Minhaj Nur Alam, Rikiya Yamashita, Vignav Ramesh, Tejas Prabhune, Jennifer I. Lim, R. V. P. Chan, Joelle Hallak, Theodore Leng, Daniel Rubin

CL based pretraining with NST significantly improves DL classification performance, helps the model generalize well (transferable from EyePACS to UIC data), and allows training with small, annotated datasets, therefore reducing ground truth annotation burden of the clinicians.

Contrastive Learning Style Transfer

A Deep-learning Approach for Prognosis of Age-Related Macular Degeneration Disease using SD-OCT Imaging Biomarkers

no code implementations • 27 Feb 2019 • Imon Banerjee, Luis de Sisternes, Joelle Hallak, Theodore Leng, Aaron Osborne, Mary Durbin, Daniel Rubin

We propose a hybrid sequential deep learning model to predict the risk of AMD progression in non-exudative AMD eyes at multiple timepoints, starting from short-term progression (3-months) up to long-term progression (21-months).