Joint Progression Modeling (JPM): A Probabilistic Framework for Mixed-Pathology Progression

2025-12-03

AI Summary

Inferring true temporal progression order in neurodegenerative diseases with mixed pathologies (e.g., comorbid Alzheimer’s disease [AD] and vascular dementia [VaD]) remains challenging from cross-sectional data alone. Method: We propose the Joint Progression Model (JPM), which models individual disease trajectories as partial-order rankings and introduces a joint progression prior grounded in probabilistic ranking frameworks—specifically Mallows and Plackett–Luce models—to enable simultaneous, interpretable, and statistically rigorous modeling and disentanglement of multiple pathological trajectories. Contribution/Results: JPM is systematically validated for calibration, separability, and sharpness. On synthetic data, it achieves a 21% improvement in ranking accuracy over SA-EBM. Applied to real-world NACC data, JPM infers AD–VaD comorbid progression patterns highly consistent with clinical literature. This work establishes a novel unsupervised temporal modeling paradigm for mixed neuropathologies.

Technology Category

• Reasoning under Uncertainty: Relational Probabilistic Models
• Machine Learning: Probabilistic Circuits and Graphical Models
• Cognitive Modeling & Cognitive Systems: Other Foundations of Cognitive Modeling & Systems

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Models for Web evolution

Abstract

Event-based models (EBMs) infer disease progression from cross-sectional data, and standard EBMs assume a single underlying disease per individual. In contrast, mixed pathologies are common in neurodegeneration. We introduce the Joint Progression Model (JPM), a probabilistic framework that treats single-disease trajectories as partial rankings and builds a prior over joint progressions. We study several JPM variants (Pairwise, Bradley-Terry, Plackett-Luce, and Mallows) and analyze three properties: (i) calibration -- whether lower model energy predicts smaller distance to the ground truth ordering; (ii) separation -- the degree to which sampled rankings are distinguishable from random permutations; and (iii) sharpness -- the stability of sampled aggregate rankings. All variants are calibrated, and all achieve near-perfect separation; sharpness varies by variant and is well-predicted by simple features of the input partial rankings (number and length of rankings, conflict, and overlap). In synthetic experiments, JPM improves ordering accuracy by roughly 21 percent over a strong EBM baseline (SA-EBM) that treats the joint disease as a single condition. Finally, using NACC, we find that the Mallows variant of JPM and the baseline model (SA-EBM) have results that are more consistent with prior literature on the possible disease progression of the mixed pathology of AD and VaD.

Problem

• Models mixed pathology progression in neurodegeneration
• Improves ordering accuracy over single-disease baseline models
• Evaluates variants on calibration, separation, and sharpness properties

Innovation

• Probabilistic framework models mixed pathology progression
• Treats single-disease trajectories as partial rankings
• Improves ordering accuracy over baseline by 21%