SSplain: Sparse and Smooth Explainer for Retinopathy of Prematurity Classification

2025-12-08

AI Summary

Existing ROP classification explainers struggle to generate pixel-level explanations that simultaneously satisfy sparsity and smoothness constraints, resulting in limited clinical interpretability and credibility. To address this, we propose SSplain—a novel explanation method that, for the first time, jointly enforces both sparsity and smoothness structural priors during explanation generation, thereby enhancing explanation naturalness and clinical consistency. SSplain formulates explanation generation as a constrained optimization problem incorporating combinatorial constraints and solves it efficiently using the Alternating Direction Method of Multipliers (ADMM) framework. Extensive experiments on multiple public ROP datasets demonstrate that SSplain significantly outperforms state-of-the-art explanation methods across key metrics—including posterior accuracy, explanation smoothness, and alignment with clinically relevant features—while exhibiting strong generalization capability.

Technology Category

• Computer Vision: Interpretability, Explainability, and Transparency
• Machine Learning: Transparent, Interpretable, Explainable ML
• Search and Optimization: Non-convex Optimization

Application Category

• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Security and Privacy: Data transparency and provenance
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Neural networks are frequently used in medical diagnosis. However, due to their black-box nature, model explainers are used to help clinicians understand better and trust model outputs. This paper introduces an explainer method for classifying Retinopathy of Prematurity (ROP) from fundus images. Previous methods fail to generate explanations that preserve input image structures such as smoothness and sparsity. We introduce Sparse and Smooth Explainer (SSplain), a method that generates pixel-wise explanations while preserving image structures by enforcing smoothness and sparsity. This results in realistic explanations to enhance the understanding of the given black-box model. To achieve this goal, we define an optimization problem with combinatorial constraints and solve it using the Alternating Direction Method of Multipliers (ADMM). Experimental results show that SSplain outperforms commonly used explainers in terms of both post-hoc accuracy and smoothness analyses. Additionally, SSplain identifies features that are consistent with domain-understandable features that clinicians consider as discriminative factors for ROP. We also show SSplain's generalization by applying it to additional publicly available datasets. Code is available at https://github.com/neu-spiral/SSplain.

Problem

• Generates pixel-wise explanations for ROP classification
• Preserves image smoothness and sparsity in explanations
• Enhances clinician trust by identifying clinically relevant features

Innovation

• Generates pixel-wise explanations preserving image smoothness and sparsity
• Solves optimization with combinatorial constraints using ADMM method
• Outperforms existing explainers in accuracy and smoothness analyses