Breast Ultrasound Tumor Generation via Mask Generator and Text-Guided Network:A Clinically Controllable Framework with Downstream Evaluation

2025-07-10

AI Summary

To address the critical challenge of scarce annotated breast ultrasound (BUS) images limiting deep learning model robustness, this paper proposes a clinically controllable, fine-grained generative framework. First, a semantic-curvature mask generator is designed to synthesize structurally diverse and anatomically plausible tumor masks by incorporating clinical prior knowledge. Second, a text-guided conditional generative adversarial network is developed to synthesize realistic BUS images conditioned on clinical descriptors (e.g., shape, echogenicity, margin). The framework enables explicit, interpretable modulation of tumor characteristics, yielding synthetically generated images with high fidelity and morphological diversity. Experiments across six public BUS datasets demonstrate substantial performance gains in downstream classification and segmentation tasks when leveraging synthetic data. A visual Turing test achieves a 92.3% pass rate, confirming clinical credibility. This work establishes a novel, interpretable, and controllable paradigm for few-shot medical image generation.

Technology Category

• Natural Language Processing: Generation
• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Machine Learning: Deep Generative Models & Autoencoders

Application Category

• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Web Mining and Content Analysis: Web data generation and simulation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

The development of robust deep learning models for breast ultrasound (BUS) image analysis is significantly constrained by the scarcity of expert-annotated data. To address this limitation, we propose a clinically controllable generative framework for synthesizing BUS images. This framework integrates clinical descriptions with structural masks to generate tumors, enabling fine-grained control over tumor characteristics such as morphology, echogencity, and shape. Furthermore, we design a semantic-curvature mask generator, which synthesizes structurally diverse tumor masks guided by clinical priors. During inference, synthetic tumor masks serve as input to the generative framework, producing highly personalized synthetic BUS images with tumors that reflect real-world morphological diversity. Quantitative evaluations on six public BUS datasets demonstrate the significant clinical utility of our synthetic images, showing their effectiveness in enhancing downstream breast cancer diagnosis tasks. Furthermore, visual Turing tests conducted by experienced sonographers confirm the realism of the generated images, indicating the framework's potential to support broader clinical applications.

Problem

• Generating synthetic breast ultrasound images with controlled tumor characteristics
• Overcoming scarcity of expert-annotated data for deep learning models
• Enhancing breast cancer diagnosis via realistic synthetic image generation

Innovation

• Generative framework integrates clinical descriptions and masks
• Semantic-curvature mask generator creates diverse tumor masks
• Synthetic images enhance breast cancer diagnosis tasks