Scholarlink

Precise segmentation of brain tumor in multimodal MRI images is a complex challenge in medical image applications. This study introduced an effective deep learning method to perform 3D segmentation of multimodal MRI images by enhancing the U-Net based structure through the incorporation of residual blocks and the module of Convolutional Block Attention (CBAM) for achieving better segmentation. The residual connections in the encoder path facilitated a deeper feature extraction process by mitigating the vanishing gradient issue. In this regard, the CBAM permits the network to adaptively refine the feature maps by explicitly targeting the most informative spatial and channel-wise components. The BraTS 2020 dataset has been used for training and evaluation. This model performs well according to different metrics compared to other existing methods. Ablation studies confirm the impact of the CBAM module in boosting overall performance.

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

A Novel Approach for Brain Tumor Segmentation Using 3D U-Net with CBAM Module

  • 章节

  • 发表日期：

    2026年

查看全文
  • Khaoula Echine,
  • Aziz Darouichi

摘要

Precise segmentation of brain tumor in multimodal MRI images is a complex challenge in medical image applications. This study introduced an effective deep learning method to perform 3D segmentation of multimodal MRI images by enhancing the U-Net based structure through the incorporation of residual blocks and the module of Convolutional Block Attention (CBAM) for achieving better segmentation. The residual connections in the encoder path facilitated a deeper feature extraction process by mitigating the vanishing gradient issue. In this regard, the CBAM permits the network to adaptively refine the feature maps by explicitly targeting the most informative spatial and channel-wise components. The BraTS 2020 dataset has been used for training and evaluation. This model performs well according to different metrics compared to other existing methods. Ablation studies confirm the impact of the CBAM module in boosting overall performance.