对比四类模型在1.1万张骨盆MRI上的分割表现，注意力网络效果最佳。
Performance Analysis of Deep Learning Models for Femur Segmentation in MRI Scan
- 用U-Net、注意力机制和KAN改进卷积结构，引入SAM2进行对比
- 最高Dice系数达0.954，注意力U-Net整体最优，U-KAN小区域更准
- 适合医学影像分割研究者参考，尤其关注小目标精确分割场景
卷积神经网络如U-Net在医学图像分割中表现优异，注意力机制与KAN可增强特征提取能力，Meta的SAM 2则基于视觉变压器实现无需微调的提示式分割。然而，这些模型在数据有限时易存在偏差，影响泛化性能。本研究系统评估并比较了三种基于CNN的模型（U-Net、Attention U-Net、U-KAN）和一种基于Transformer的模型（SAM 2），用于磁共振成像（MRI）中股骨结构的分割。数据集包含11,164例带精细股骨标注的MRI扫描图像。性能通过骰子相似系数（Dice Similarity Coefficient）评估，结果范围为0.932至0.954。Attention U-Net总体表现最佳，而U-KAN在感兴趣区域较小的解剖部位表现出更优性能，得益于其更强的学习能力，提升了分割精度。
原文摘要 · Abstract (English)
Convolutional neural networks like U-Net excel in medical image segmentation, while attention mechanisms and KAN enhance feature extraction. Meta's SAM 2 uses Vision Transformers for prompt-based segmentation without fine-tuning. However, biases in these models impact generalization with limited data. In this study, we systematically evaluate and compare the performance of three CNN-based models, i.e., U-Net, Attention U-Net, and U-KAN, and one transformer-based model, i.e., SAM 2 for segmenting femur bone structures in MRI scan. The dataset comprises 11,164 MRI scans with detailed annotations of femoral regions. Performance is assessed using the Dice Similarity Coefficient, which ranges from 0.932 to 0.954. Attention U-Net achieves the highest overall scores, while U-KAN demonstrated superior performance in anatomical regions with a smaller region of interest, leveraging its enhanced learning capacity to improve segmentation accuracy.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。