arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2212.04497cs.CV
收藏

UNETR++:探索高效且精确的3D医学图像分割

UNETR++: Delving into Efficient and Accurate 3D Medical Image Segmentation

Mohamed bin Zayed University of AI(穆罕默德·本·扎耶德人工智能大学);Australian National University(澳大利亚国立大学);University of California, Merced(加州大学默塞德分校);Yonsei University(延世大学);Google Research(谷歌研究院);Linköping University(林雪平大学)

Abdelrahman Shaker, Muhammad Maaz, Hanoona Rasheed, Salman Khan, Ming-Hsuan Yang, Fahad Shahbaz Khan

AI总结:

本文提出UNETR++,通过高效配对注意力(EPA)模块以线性复杂度学习空间与通道判别特征,在五个3D医学分割基准上兼顾高精度与高效率,并在Synapse上以87.2% Dice创下新SOTA,同时参数量和FLOPs减少超71%。

摘要:

由于Transformer模型的成功,近期工作研究了其在3D医学分割任务中的适用性。在Transformer模型中,自注意力机制是主要构建模块之一,致力于捕获长程依赖关系。然而,自注意力操作具有二次复杂度,这被证明是一个计算瓶颈,尤其是在体积医学成像中,其输入是包含众多切片的3D数据。在本文中,我们提出了一种名为UNETR++的3D医学图像分割方法,该方法既能提供高质量的分割掩码,又在参数量、计算成本和推理速度方面具有高效性。我们设计的核心是引入了一种新颖的高效配对注意力(EPA)模块,该模块利用基于空间注意力和通道注意力的两个相互依赖分支,高效地学习空间和通道维度的判别性特征。我们的空间注意力公式是高效的,相对于输入序列长度具有线性复杂度。为了实现空间聚焦分支和通道聚焦分支之间的通信,我们共享查询和键映射函数的权重,这提供了互补性优势(配对注意力),同时也减少了整体网络参数。我们在五个基准数据集(Synapse、BTCV、ACDC、BRaTs和Decathlon-Lung)上的广泛评估揭示了我们贡献在效率和准确性方面的有效性。在Synapse上,我们的UNETR++以87.2%的Dice Score创造了新的最先进水平,同时与文献中的最佳方法相比,在参数量和FLOPs方面均减少了超过71%,显著高效。代码:https://github.com/Amshaker/unetr_plus_plus。

英文摘要:

Owing to the success of transformer models, recent works study their applicability in 3D medical segmentation tasks. Within the transformer models, the self-attention mechanism is one of the main building blocks that strives to capture long-range dependencies. However, the self-attention operation has quadratic complexity which proves to be a computational bottleneck, especially in volumetric medical imaging, where the inputs are 3D with numerous slices. In this paper, we propose a 3D medical image segmentation approach, named UNETR++, that offers both high-quality segmentation masks as well as efficiency in terms of parameters, compute cost, and inference speed. The core of our design is the introduction of a novel efficient paired attention (EPA) block that efficiently learns spatial and channel-wise discriminative features using a pair of inter-dependent branches based on spatial and channel attention. Our spatial attention formulation is efficient having linear complexity with respect to the input sequence length. To enable communication between spatial and channel-focused branches, we share the weights of query and key mapping functions that provide a complimentary benefit (paired attention), while also reducing the overall network parameters. Our extensive evaluations on five benchmarks, Synapse, BTCV, ACDC, BRaTs, and Decathlon-Lung, reveal the effectiveness of our contributions in terms of both efficiency and accuracy. On Synapse, our UNETR++ sets a new state-of-the-art with a Dice Score of 87.2%, while being significantly efficient with a reduction of over 71% in terms of both parameters and FLOPs, compared to the best method in the literature. Code: https://github.com/Amshaker/unetr_plus_plus.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑