Scholarlink

Accurate segmentation of brain metastasis is essential for treatment planning and surgical procedures. Still, it is difficult due to the heterogeneous nature and complex spatial relationships of tumor regions. The objective of this study is to develop an automated segmentation model that labels an MRI image into the background+resection cavity (BC+RC), non-enhancing tumor core (NETC), surrounding non-enhancing FLAIR hyperintensity (SNFH), and enhancing tumor (ET) regions. For this purpose, we proposed a deep learning model, SECBAMUNet3D, that integrates the standard U-Net architecture with CBAM and Squeeze-and-Excitation blocks, as well as residual dense blocks. It is trained on the BraTS dataset using a combined loss function and achieves 71.2% Dice coefficient, 7.0 mm Hausdorff distance (95th percentile), and 1.8 mm average symmetric surface distance, outperforming all SOTA models. These results demonstrate the robustness and practicality of the proposed model in real-world clinical settings.

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

Cross-Modal Attention and Residual Dense Learning Approach for Multimodal Brain Tumor Segmentation Using 3D MRI Scans

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Muhammad Attique Khan,
  • Najib Ben Aoun,
  • Muhammad John Abbas,
  • Ghassen Ben Brahim

摘要

Accurate segmentation of brain metastasis is essential for treatment planning and surgical procedures. Still, it is difficult due to the heterogeneous nature and complex spatial relationships of tumor regions. The objective of this study is to develop an automated segmentation model that labels an MRI image into the background+resection cavity (BC+RC), non-enhancing tumor core (NETC), surrounding non-enhancing FLAIR hyperintensity (SNFH), and enhancing tumor (ET) regions. For this purpose, we proposed a deep learning model, SECBAMUNet3D, that integrates the standard U-Net architecture with CBAM and Squeeze-and-Excitation blocks, as well as residual dense blocks. It is trained on the BraTS dataset using a combined loss function and achieves 71.2% Dice coefficient, 7.0 mm Hausdorff distance (95th percentile), and 1.8 mm average symmetric surface distance, outperforming all SOTA models. These results demonstrate the robustness and practicality of the proposed model in real-world clinical settings.