Directed Ordinal Diffusion Regularization for Progression-Aware Diabetic Retinopathy Grading

TL;DR

This paper introduces Directed Ordinal Diffusion Regularization (D-ODR), a novel method that models diabetic retinopathy progression as a directed flow in feature space, improving grading accuracy by enforcing biologically plausible disease evolution.

Contribution

D-ODR explicitly encodes disease progression as a directed graph and applies multi-scale diffusion to prevent reverse transitions, aligning feature learning with clinical progression.

Findings

D-ODR outperforms existing methods in grading accuracy.

The method enforces forward disease progression in feature space.

Results demonstrate improved clinical reliability of severity assessment.

Abstract

Diabetic Retinopathy (DR) progresses as a continuous and irreversible deterioration of the retina, following a well-defined clinical trajectory from mild to severe stages. However, most existing ordinal regression approaches model DR severity as a set of static, symmetric ranks, capturing relative order while ignoring the inherent unidirectional nature of disease progression. As a result, the learned feature representations may violate biological plausibility, allowing implausible proximity between non-consecutive stages or even reverse transitions. To bridge this gap, we propose Directed Ordinal Diffusion Regularization (D-ODR), which explicitly models the feature space as a directed flow by constructing a progression-constrained directed graph that strictly enforces forward disease evolution. By performing multi-scale diffusion on this directed structure, D-ODR imposes penalties on score inversions along valid progression paths, thereby effectively preventing the model from learning biologically inconsistent reverse transitions. This mechanism aligns the feature representation with the natural trajectory of DR worsening. Extensive experiments demonstrate that D-ODR yields superior grading performance compared to state-of-the-art ordinal regression and DR-specific grading methods, offering a more clinically reliable assessment of disease severity. Our code is available on https://github.com/HovChen/D-ODR.