Ultrasound Image Synthesis Using Generative AI for Lung Ultrasound Detection

2025-01-10

AI Summary

To address the challenges of data scarcity and severe class imbalance in lung ultrasound (LUS) imaging—which hinder accurate detection of rare pathologies such as large pulmonary consolidations—this paper proposes DiffUltra, the first high-fidelity, lesion-diversity-controllable generative method tailored for LUS. Its core innovation is a Lesion-Anatomy Bank that integrates clinical anatomical priors and empirically observed spatial lesion distributions into a conditional diffusion model, enabling anatomy- and location-aware lesion synthesis. Evaluated on consolidation detection, DiffUltra improves overall average precision (AP) by 5.6%; notably, for large consolidations—constituting only 10% of cases—it achieves a 25% AP gain, substantially mitigating long-tail distribution effects. This work establishes a novel paradigm for few-shot medical image generation and downstream diagnostic tasks.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Generation

Application Category

• Search and Retrieval-Augmented AI: Large language models for search
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI

Abstract

Developing reliable healthcare AI models requires training with representative and diverse data. In imbalanced datasets, model performance tends to plateau on the more prevalent classes while remaining low on less common cases. To overcome this limitation, we propose DiffUltra, the first generative AI technique capable of synthesizing realistic Lung Ultrasound (LUS) images with extensive lesion variability. Specifically, we condition the generative AI by the introduced Lesion-anatomy Bank, which captures the lesion's structural and positional properties from real patient data to guide the image synthesis.We demonstrate that DiffUltra improves consolidation detection by 5.6% in AP compared to the models trained solely on real patient data. More importantly, DiffUltra increases data diversity and prevalence of rare cases, leading to a 25% AP improvement in detecting rare instances such as large lung consolidations, which make up only 10% of the dataset.

Problem

• Medical AI
• Rare Pulmonary Lesions
• Ultrasound Imaging

Innovation

• DiffUltra
• AI-generated lung ultrasound images
• Improved diagnostic accuracy for rare and large pulmonary lesions