A Sparse-Attention Deep Learning Model Integrating Heterogeneous Multimodal Features for Parkinson's Disease Severity Profiling

TL;DR

This paper introduces SAFN, an interpretable deep learning model that effectively fuses multimodal data to accurately profile Parkinson's disease severity, addressing challenges like interpretability and class imbalance.

Contribution

SAFN is a novel deep learning framework that integrates multimodal neuroimaging and clinical data using sparse attention and class-balanced loss, improving accuracy and interpretability in PD profiling.

Findings

Achieved 98% accuracy and perfect PR-AUC on PD dataset.

Effectively prioritized clinical assessments in decision-making.

Outperformed existing machine learning and deep learning baselines.

Abstract

Characterising the heterogeneous presentation of Parkinson's disease (PD) requires integrating biological and clinical markers within a unified predictive framework. While multimodal data provide complementary information, many existing computational models struggle with interpretability, class imbalance, or effective fusion of high-dimensional imaging and tabular clinical features. To address these limitations, we propose the Class-Weighted Sparse-Attention Fusion Network (SAFN), an interpretable deep learning framework for robust multimodal profiling. SAFN integrates MRI cortical thickness, MRI volumetric measures, clinical assessments, and demographic variables using modality-specific encoders and a symmetric cross-attention mechanism that captures nonlinear interactions between imaging and clinical representations. A sparsity-constrained attention-gating fusion layer dynamically prioritises informative modalities, while a class-balanced focal loss (beta = 0.999, gamma = 1.5) mitigates dataset imbalance without synthetic oversampling. Evaluated on 703 participants (570 PD, 133 healthy controls) from the Parkinson's Progression Markers Initiative using subject-wise five-fold cross-validation, SAFN achieves an accuracy of 0.98 plus or minus 0.02 and a PR-AUC of 1.00 plus or minus 0.00, outperforming established machine learning and deep learning baselines. Interpretability analysis shows a clinically coherent decision process, with approximately 60 percent of predictive weight assigned to clinical assessments, consistent with Movement Disorder Society diagnostic principles. SAFN provides a reproducible and transparent multimodal modelling paradigm for computational profiling of neurodegenerative disease.