Query Nearby：偏移调整的 Mask2Former 增强小器官分割
Query Nearby: Offset-Adjusted Mask2Former enhances small-organ segmentation
AI总结：
针对医学图像中小器官分割效果差、transformer 训练成本高的问题，本文提出带偏移调整、粗略位置特征和 FCN 辅助头的 Mask2Former，并在 HaNSeg 与 SegRap2023 上达到 SOTA。
摘要：
医学分割在放射治疗和手术引导等临床应用中发挥着重要作用，但获得临床可接受的结果十分困难。近年来，随着类 transformer 模型的成功应用，例如将注意力机制与 CNN 相结合，该领域已取得进展。特别是，基于 transformer 的分割模型能够更有效地提取全局信息，弥补 CNN 模块关注局部特征的缺陷。然而，使用 transformer 架构并不容易，因为训练基于 transformer 的模型可能需要大量资源。此外，由于医学领域的鲜明特征，尤其是在遇到区域紧凑的中型和小型器官时，其结果往往不能令人满意。例如，直接使用 ViT 分割医学图像只能得到低于 50% 的 DSC，远低于 80% 的临床可接受分数。本文中，我们使用带有可变形注意力的 Mask2Former 来减少计算，并提出偏移调整策略，以在注意力权重计算过程中促使采样点位于同一器官内，从而更好地整合紧凑的前景信息。此外，我们利用 Mask2Former 中的第 4 个特征图提供器官的粗略位置，并采用基于 FCN 的辅助头，通过 Dice loss 帮助更快地训练 Mask2Former。我们表明，我们的模型在 HaNSeg 和 SegRap2023 数据集上取得了 SOTA（State-of-the-Art，最先进）性能，尤其是在中型和小型器官上。我们的代码可在链接 https://github.com/earis/Offsetadjustment_Background-location_Decoder_Mask2former 获取。
英文摘要：
Medical segmentation plays an important role in clinical applications like radiation therapy and surgical guidance, but acquiring clinically acceptable results is difficult. In recent years, progress has been witnessed with the success of utilizing transformer-like models, such as combining the attention mechanism with CNN. In particular, transformer-based segmentation models can extract global information more effectively, compensating for the drawbacks of CNN modules that focus on local features. However, utilizing transformer architecture is not easy, because training transformer-based models can be resource-demanding. Moreover, due to the distinct characteristics in the medical field, especially when encountering mid-sized and small organs with compact regions, their results often seem unsatisfactory. For example, using ViT to segment medical images directly only gives a DSC of less than 50\%, which is far lower than the clinically acceptable score of 80\%. In this paper, we used Mask2Former with deformable attention to reduce computation and proposed offset adjustment strategies to encourage sampling points within the same organs during attention weights computation, thereby integrating compact foreground information better. Additionally, we utilized the 4th feature map in Mask2Former to provide a coarse location of organs, and employed an FCN-based auxiliary head to help train Mask2Former more quickly using Dice loss. We show that our model achieves SOTA (State-of-the-Art) performance on the HaNSeg and SegRap2023 datasets, especially on mid-sized and small organs.Our code is available at link https://github.com/earis/Offsetadjustment\_Background-location\_Decoder\_Mask2former.
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。
