Scholarlink

<b>Purpose:</b> <p>In this paper, we propose a novel generative model to produce high-quality SAH samples, enhancing SAH CT detection performance in imbalanced datasets. Previous methods, such as cost-sensitive learning and previous diffusion models, suffer from overfitting or noise-induced distortion, limiting their effectiveness. Accurate SAH sample generation is crucial for better detection.</p> <b>Methods:</b> <p>We propose the Worley–Perlin Diffusion Model (WPDM), leveraging Worley–Perlin noise to synthesize diverse, high-quality SAH images. WPDM addresses limitations of Gaussian noise (homogeneity) and Simplex noise (distortion), enhancing robustness for generating SAH images. Additionally, <InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(\hbox {WPDM}_{\text {Fast}}\)</EquationSource> <EquationSource Format="MATHML"><math> <msub> <mtext>WPDM</mtext> <mtext>Fast</mtext> </msub> </math></EquationSource> </InlineEquation> optimizes generation speed without compromising quality.</p> <b>Results:</b> <p>WPDM effectively improved classification accuracy in datasets with varying imbalance ratios. Notably, a classifier trained with WPDM-generated samples achieved an F1-score of 0.857 on a 1:36 imbalance ratio, surpassing the state of the art by 2.3 percentage points.</p> <b>Conclusion:</b> <p>WPDM overcomes the limitations of Gaussian and Simplex noise-based models, generating high-quality, realistic SAH images. It significantly enhances classification performance in imbalanced settings, providing a robust solution for SAH CT detection.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Synthetic data generation with Worley–Perlin diffusion for robust subarachnoid hemorrhage detection in imbalanced CT Datasets

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年9月2日

查看全文
  • Zhongyang Lu,
  • Tao Hu,
  • Masahiro Oda,
  • Yutaro Fuse,
  • Ryuta Saito,
  • Masahiro Jinzaki,
  • Kensaku Mori

摘要

Purpose:

In this paper, we propose a novel generative model to produce high-quality SAH samples, enhancing SAH CT detection performance in imbalanced datasets. Previous methods, such as cost-sensitive learning and previous diffusion models, suffer from overfitting or noise-induced distortion, limiting their effectiveness. Accurate SAH sample generation is crucial for better detection.

Methods:

We propose the Worley–Perlin Diffusion Model (WPDM), leveraging Worley–Perlin noise to synthesize diverse, high-quality SAH images. WPDM addresses limitations of Gaussian noise (homogeneity) and Simplex noise (distortion), enhancing robustness for generating SAH images. Additionally, \(\hbox {WPDM}_{\text {Fast}}\) WPDM Fast optimizes generation speed without compromising quality.

Results:

WPDM effectively improved classification accuracy in datasets with varying imbalance ratios. Notably, a classifier trained with WPDM-generated samples achieved an F1-score of 0.857 on a 1:36 imbalance ratio, surpassing the state of the art by 2.3 percentage points.

Conclusion:

WPDM overcomes the limitations of Gaussian and Simplex noise-based models, generating high-quality, realistic SAH images. It significantly enhances classification performance in imbalanced settings, providing a robust solution for SAH CT detection.