BioDiffusion: A Versatile Diffusion Model for Biomedical Signal Synthesis

TL;DR

BioDiffusion is a diffusion-based probabilistic model designed to generate high-quality, multivariate biomedical signals, effectively addressing data scarcity, noise, and imbalance issues in biomedical machine learning tasks.

Contribution

The paper introduces BioDiffusion, a novel diffusion model tailored for biomedical signal synthesis, outperforming existing models in quality and versatility.

Findings

BioDiffusion produces high-fidelity, non-stationary signals.

It improves machine learning accuracy with synthesized data.

Outperforms current state-of-the-art time-series generative models.

Abstract

Machine learning tasks involving biomedical signals frequently grapple with issues such as limited data availability, imbalanced datasets, labeling complexities, and the interference of measurement noise. These challenges often hinder the optimal training of machine learning algorithms. Addressing these concerns, we introduce BioDiffusion, a diffusion-based probabilistic model optimized for the synthesis of multivariate biomedical signals. BioDiffusion demonstrates excellence in producing high-fidelity, non-stationary, multivariate signals for a range of tasks including unconditional, label-conditional, and signal-conditional generation. Leveraging these synthesized signals offers a notable solution to the aforementioned challenges. Our research encompasses both qualitative and quantitative assessments of the synthesized data quality, underscoring its capacity to bolster accuracy in machine learning tasks tied to biomedical signals. Furthermore, when juxtaposed with current leading time-series generative models, empirical evidence suggests that BioDiffusion outperforms them in biomedical signal generation quality.