H-LDM: Hierarchical Latent Diffusion Models for Controllable and Interpretable PCG Synthesis from Clinical Metadata

TL;DR

H-LDM is a hierarchical latent diffusion model that generates controllable, clinically accurate PCG signals from metadata, improving data augmentation for cardiac diagnosis and enabling interpretability.

Contribution

It introduces a multi-scale VAE and a hierarchical pipeline with a novel Medical Attention module for physiologically-disentangled, controllable PCG synthesis from clinical metadata.

Findings

Achieves state-of-the-art Fréchet Audio Distance of 9.7

92% attribute disentanglement score

87.1% clinical validity confirmed by cardiologists

Abstract

Phonocardiogram (PCG) analysis is vital for cardiovascular disease diagnosis, yet the scarcity of labeled pathological data hinders the capability of AI systems. To bridge this, we introduce H-LDM, a Hierarchical Latent Diffusion Model for generating clinically accurate and controllable PCG signals from structured metadata. Our approach features: (1) a multi-scale VAE that learns a physiologically-disentangled latent space, separating rhythm, heart sounds, and murmurs; (2) a hierarchical text-to-biosignal pipeline that leverages rich clinical metadata for fine-grained control over 17 distinct conditions; and (3) an interpretable diffusion process guided by a novel Medical Attention module. Experiments on the PhysioNet CirCor dataset demonstrate state-of-the-art performance, achieving a Fr\'echet Audio Distance of 9.7, a 92% attribute disentanglement score, and 87.1% clinical validity confirmed by cardiologists. Augmenting diagnostic models with our synthetic data improves the accuracy of rare disease classification by 11.3\%. H-LDM establishes a new direction for data augmentation in cardiac diagnostics, bridging data scarcity with interpretable clinical insights.