Forecasting-based biomedical time-series data synthesis for open data and robust AI

Clinical Snapshot

Key Findings

• P Value: Not reported

• Effect Size: 3.71% performance improvement with data augmentation

• Primary Outcome: Sleep-stage classification accuracy using synthetic data

• Nnt Or Sensitivity: 91.00% synthetic-only accuracy versus real-data baseline

• Confidence Interval: Not reported

Clinical Application

Requires computational infrastructure and expertise in forecasting models; implementation complexity unclear from abstract Could support Australian sleep research initiatives and TGA-regulated medical device development while addressing privacy requirements under Australian Privacy Principles Sleep medicine, neurophysiology, and electrophysiology research communities requiring large datasets for AI development

Abstract

The limited data availability due to strict privacy regulations and significant resource demands severely constrains biomedical time-series AI development, which creates a critical gap between data requirements and accessibility. Synthetic data generation presents a promising solution by producing artificial datasets that maintain the statistical properties of real biomedical time-series data without compromising patient confidentiality. While GANs, VAEs, and diffusion models capture global data distributions, forecasting models offer inductive biases tailored for sequential dynamics. We propose a framework for synthetic biomedical time-series data generation based on recent forecasting models that accurately replicates complex electrophysiological signals such as EEG and EMG with high fidelity. These synthetic datasets can be freely shared for open AI development and consistently improve downstream model performance. Numerical results on sleep-stage classification show up to a 3.71% performance gain with augmentation and a 91.00% synthetic-only accuracy that surpasses the real-data-only baseline.