HyperNSDE: Personalized Neural SDEs for Joint Static-Longitudinal Clinical Data Generation

2026-10-05

AI Summary

This study addresses the challenge of jointly modeling static covariates, irregular longitudinal trajectories, and observation times under medical data scarcity and privacy constraints. To this end, we propose a continuous-time generative model that conditions neural stochastic differential equations on static features via hypernetworks. Notably, we introduce a pioneering guidance mechanism that eliminates the need for trajectory encoders, while simultaneously modeling state-dependent observation intensity processes. Furthermore, signature kernels are incorporated to enable non-adversarial training. Experimental results demonstrate that the proposed model significantly improves observation time fidelity on both simulated and real-world datasets, yielding competitive synthetic data quality. Overall, this work establishes a novel paradigm for clinical data generation.

Abstract

Synthetic patient data generation is a promising solution to the dual challenge of data scarcity and privacy constraints in healthcare machine learning. Realistic synthesis of patient-level clinical data requires jointly modeling heterogeneous static covariates, irregularly sampled longitudinal trajectories, and informative observation times - three tightly coupled components in practice yet rarely addressed together. We propose HyperNSDE, a continuous-time generative model that conditions a latent Neural SDE on static patient representations through a hypernetwork, allowing baseline characteristics to shape trajectory evolution beyond the initial condition without requiring a trajectory encoder, while stochastic latent dynamics capture realistic variability in generated paths. Observation times are modeled jointly through a latent-state-dependent intensity process, and training on irregular stochastic paths is stabilized via a deterministic-stochastic path decomposition with a non-adversarial signature-kernel objective. Experiments on simulated and real clinical datasets show improved observation-time fidelity and competitive performance, while matched-grid analyses reveal that forecasting and correlation metrics are affected by observation-grid regularity and trajectory smoothness.

Problem

• synthetic patient data generation
• static covariates
• longitudinal trajectories
• observation times
• clinical data

Innovation

• Neural SDE
• Hypernetwork
• Clinical Data Generation
• Irregularly Sampled Longitudinal Data
• Signature-Kernel Objective