Conditional Evidence Reconstruction and Decomposition for Interpretable Multimodal Diagnosis

2026-04-18

AI Summary

Multimodal disease diagnosis often suffers significant performance degradation under missing modalities and lacks interpretability. To address these challenges, this work proposes a novel approach that conditionally reconstructs missing modalities based on individual-specific features and decomposes diagnostic evidence at the logit level into shared and modality-specific components, enabling robust and interpretable fusion decisions. Evaluated on the ADNI dataset, the proposed method substantially outperforms existing techniques, maintaining high diagnostic accuracy even with incomplete modalities while providing evidence attributions that align with clinical understanding—thus achieving both robustness and interpretability.

Technology Category

• Computer Vision: Multi-modal Vision
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Machine Learning: Multimodal Learning

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs

Abstract

Neurobiological and neurodegenerative diseases are inherently multifactorial, arising from coupled influences spanning genetic susceptibility, brain alterations, and environmental and behavioral factors. Multimodal modeling has therefore been increasingly adopted for disease diagnosis by integrating complementary evidence across data sources. However, in both large-scale cohorts and real-world clinical workflows, modality coverage is often incomplete, making many multimodal models brittle when one or more modalities are unavailable. Existing approaches to incomplete multimodal diagnosis typically rely on group-wise or static priors, which may fail to capture subject-specific cross-modal dependencies; moreover, many models provide limited interpretability into which evidence sources drive the final decision. To address these limitations, we propose Conditional Evidence Reconstruction and Decomposition (CERD), a framework for interpretable multimodal diagnosis with incomplete modalities. CERD first reconstructs missing modality representations conditioned on each subject's observed inputs, then decomposes diagnostic evidence into shared cross-modal corroboration and modality-specific cues via logit-level attribution. Experiments on the Alzheimer's Disease Neuroimaging Initiative (ADNI) demonstrate that CERD outperforms competitive baselines under incomplete-modality settings while producing structured and clinically aligned evidence attributions for trustworthy decision support.

Problem

• incomplete multimodal diagnosis
• cross-modal dependencies
• interpretability
• evidence attribution
• multimodal modeling

Innovation

• Conditional Evidence Reconstruction
• Multimodal Diagnosis
• Incomplete Modalities
• Evidence Decomposition
• Interpretable AI