arXiv 2501.028822025-01-06cs.CV

PARF-Net：将逐像素自适应感受野集成到用于医学图像分割的混合 Transformer-CNN 网络中

PARF-Net: integrating pixel-wise adaptive receptive fields into hybrid Transformer-CNN network for medical image segmentation

School of Information Engineering, Ningxia University（宁夏大学信息工程学院）

Xu Ma, Mengsheng Chen, Junhui Zhang, Lijuan Song, Fang Du, Zhenhua Yu

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对混合 Transformer-CNN 固定感受野及依赖融合不足的问题，本文提出 PARF-Net，利用逐像素自适应卷积和轻量混合模块提升医学图像分割性能。

摘要：

卷积神经网络（CNNs）擅长局部特征提取，而 Transformers 在处理全局语义信息方面更具优势。通过结合两者的优势，混合 Transformer-CNN 网络已成为医学图像分割任务中的主要架构。然而，由于卷积的感受野固定，现有混合方法仍存在局部语义特征学习不足的问题，并且在有效整合局部与长程依赖方面也存在欠缺。为解决这些问题，我们开发了一种新方法 PARF-Net，将逐像素自适应感受野卷积（Conv-PARF）集成到用于医学图像分割的混合网络中。引入 Conv-PARF 是为了应对像素间语义差异，并为每个像素动态调整卷积感受野，从而提供具有区分度的特征，以将形状和尺度各异的病灶从背景中分离出来。来自 Conv-PARF 层的特征进一步由混合 Transformer-CNN 模块以轻量级方式处理，以有效捕获局部和长程依赖，从而提升分割性能。通过在 MoNuSeg、GlaS、DSB2018 和多器官 Synapse 四个广泛使用的医学图像数据集上评估 PARF-Net，我们展示了该方法相较于最先进方法的优势。例如，PARF-Net 在 Synapse 数据集上取得了 84.27% 的平均 Dice，大幅超越现有方法。

英文摘要：

Convolutional neural networks (CNNs) excel in local feature extraction while Transformers are superior in processing global semantic information. By leveraging the strengths of both, hybrid Transformer-CNN networks have become the major architectures in medical image segmentation tasks. However, existing hybrid methods still suffer deficient learning of local semantic features due to the fixed receptive fields of convolutions, and also fall short in effectively integrating local and long-range dependencies. To address these issues, we develop a new method PARF-Net to integrate convolutions of Pixel-wise Adaptive Receptive Fields (Conv-PARF) into hybrid Network for medical image segmentation. The Conv-PARF is introduced to cope with inter-pixel semantic differences and dynamically adjust convolutional receptive fields for each pixel, thus providing distinguishable features to disentangle the lesions with varying shapes and scales from the background. The features derived from the Conv-PARF layers are further processed using hybrid Transformer-CNN blocks under a lightweight manner, to effectively capture local and long-range dependencies, thus boosting the segmentation performance. By assessing PARF-Net on four widely used medical image datasets including MoNuSeg, GlaS, DSB2018 and multi-organ Synapse, we showcase the advantages of our method over the state-of-the-arts. For instance, PARF-Net achieves 84.27% mean Dice on the Synapse dataset, surpassing existing methods by a large margin.

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。