This work proposes a novel framework that integrates self-supervised learning with neural ordinary differential equations (NODEs) to effectively model and predict disease progression, specifically focusing on diabetic retinopathy.
