AWDiff: An a trous wavelet diffusion model for lung ultrasound image synthesis
This addresses data scarcity in lung ultrasound imaging for medical AI applications, but it is incremental as it builds on existing diffusion models with wavelet integration.
The paper tackled the problem of limited lung ultrasound data for machine learning by proposing AWDiff, a diffusion-based augmentation framework that preserves fine-scale diagnostic cues like B-lines, achieving lower distortion and higher perceptual quality compared to existing methods.
Lung ultrasound (LUS) is a safe and portable imaging modality, but the scarcity of data limits the development of machine learning methods for image interpretation and disease monitoring. Existing generative augmentation methods, such as Generative Adversarial Networks (GANs) and diffusion models, often lose subtle diagnostic cues due to resolution reduction, particularly B-lines and pleural irregularities. We propose A trous Wavelet Diffusion (AWDiff), a diffusion based augmentation framework that integrates the a trous wavelet transform to preserve fine-scale structures while avoiding destructive downsampling. In addition, semantic conditioning with BioMedCLIP, a vision language foundation model trained on large scale biomedical corpora, enforces alignment with clinically meaningful labels. On a LUS dataset, AWDiff achieved lower distortion and higher perceptual quality compared to existing methods, demonstrating both structural fidelity and clinical diversity.