Scholarlink

<p>Three-dimensional (3D) medical image registration has drawn substantial research attention. In comparison to traditional approaches, deep learning techniques present significant advantages in terms of speed and accuracy. However, large deformations and complex transformations pose challenges for single-modality image registration. In this study, we propose WTDL-Net, a multi-scale registration network incorporating wavelet transform. First, low-frequency sub-images generated by WT at various resolutions are used as inputs to the multi-scale registration network. Coarse-to-fine registration is achieved by analyzing image information at different resolutions. Second, the high-frequency components derived from the WT are combined to create a high-frequency infographic. This Infographic is applied to constrain multi-level registration, thereby enhancing the optimization of registration details. The proposed approach outperforms existing deep learning-based registration techniques, as shown through comprehensive quantitative and qualitative evaluations on four MR brain scan datasets.</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

WTDL-Net: medical image registration based on wavelet transform and multi-scale deep learning

  • 期刊论文

  • 发表日期：

    2025年6月26日

查看全文
  • BoHua Chu,
  • BaoJu Zhang,
  • Bo Zhang,
  • CuiPing Zhang

摘要

Three-dimensional (3D) medical image registration has drawn substantial research attention. In comparison to traditional approaches, deep learning techniques present significant advantages in terms of speed and accuracy. However, large deformations and complex transformations pose challenges for single-modality image registration. In this study, we propose WTDL-Net, a multi-scale registration network incorporating wavelet transform. First, low-frequency sub-images generated by WT at various resolutions are used as inputs to the multi-scale registration network. Coarse-to-fine registration is achieved by analyzing image information at different resolutions. Second, the high-frequency components derived from the WT are combined to create a high-frequency infographic. This Infographic is applied to constrain multi-level registration, thereby enhancing the optimization of registration details. The proposed approach outperforms existing deep learning-based registration techniques, as shown through comprehensive quantitative and qualitative evaluations on four MR brain scan datasets.