中文标题#
MLRU++：具有注意力机制的多尺度轻量级残差 UNETR++ 用于高效的三维医学图像分割
英文标题#
MLRU++: Multiscale Lightweight Residual UNETR++ with Attention for Efficient 3D Medical Image Segmentation
中文摘要#
準確且高效的醫學圖像分割由於解剖學的變異性以及對體積數據的高計算需求而具有挑戰性。最近的混合卷積神經網絡 - Transformer 架構取得了最先進的結果，但增加了顯著的複雜性。在本文中，我們提出了 MLRU++，一種多尺度輕量級殘差 UNETR++ 架構，旨在平衡分割精度和計算效率。它引入了兩個關鍵創新：一個輕量級通道和瓶頸注意力模塊（LCBAM），以最小的開銷增強上下文特徵編碼，以及解碼器中的多尺度瓶頸塊（M2B），通過多分辨率特徵聚合捕捉細粒度細節。在四個公開可用的基準數據集（Synapse、BTCV、ACDC 和 Decathlon Lung）上的實驗表明，MLRU++ 取得了最先進的性能，平均 Dice 分數分別為 87.57%（Synapse）、93.00%（ACDC）和 81.12%（Lung）。與現有領先模型相比，MLRU++ 在 Synapse 和 ACDC 上的 Dice 分數分別提高了 5.38% 和 2.12%，同時顯著減少了參數數量和計算成本。評估 LCBAM 和 M2B 的消融研究進一步證實了所提出架構組件的有效性。結果表明，MLRU++ 為 3D 醫學圖像分割任務提供了一個實用且高性能的解決方案。源代碼可在以下位置獲取：https://github.com/1027865/MLRUPP
英文摘要#
Accurate and efficient medical image segmentation is crucial but challenging due to anatomical variability and high computational demands on volumetric data. Recent hybrid CNN-Transformer architectures achieve state-of-the-art results but add significant complexity. In this paper, we propose MLRU++, a Multiscale Lightweight Residual UNETR++ architecture designed to balance segmentation accuracy and computational efficiency. It introduces two key innovations: a Lightweight Channel and Bottleneck Attention Module (LCBAM) that enhances contextual feature encoding with minimal overhead, and a Multiscale Bottleneck Block (M2B) in the decoder that captures fine-grained details via multi-resolution feature aggregation. Experiments on four publicly available benchmark datasets (Synapse, BTCV, ACDC, and Decathlon Lung) demonstrate that MLRU++ achieves state-of-the-art performance, with average Dice scores of 87.57% (Synapse), 93.00% (ACDC), and 81.12% (Lung). Compared to existing leading models, MLRU++ improves Dice scores by 5.38% and 2.12% on Synapse and ACDC, respectively, while significantly reducing parameter count and computational cost. Ablation studies evaluating LCBAM and M2B further confirm the effectiveness of the proposed architectural components. Results suggest that MLRU++ offers a practical and high-performing solution for 3D medical image segmentation tasks. Source code is available at: https://github.com/1027865/MLRUPP
PDF 获取#
抖音掃碼查看更多精彩內容