论文阅读 [TPAMI-2022] Learnable Pooling in Graph Convolutional Networks for Brain Surface Analysis

论文阅读 [TPAMI-2022] Learnable Pooling in Graph Convolutional Networks for Brain Surface Analysis

论文搜索(studyai.com)

搜索论文: Learnable Pooling in Graph Convolutional Networks for Brain Surface Analysis

搜索论文: http://www.studyai.com/search/whole-site/?q=Learnable+Pooling+in+Graph+Convolutional+Networks+for+Brain+Surface+Analysis

关键字(Keywords)

Brain; Convolution; Geometry; Task analysis; Surface treatment; Alzheimer’s disease; Learnable pooling; graph convolutional networks; brain surface analysis; alzheimer classification

机器学习; 机器视觉

年龄估计; 图卷积网络; 三维成像

摘要(Abstract)

Brain surface analysis is essential to neuroscience, however, the complex geometry of the brain cortex hinders computational methods for this task.

大脑表面分析对神经科学至关重要，然而，大脑皮层的复杂几何结构阻碍了这项任务的计算方法。.

The difficulty arises from a discrepancy between 3D imaging data, which is represented in Euclidean space, and the non-Euclidean geometry of the highly-convoluted brain surface.

困难源于以欧几里德空间表示的3D成像数据与高度卷积的大脑表面的非欧几里德几何之间的差异。.

Recent advances in machine learning have enabled the use of neural networks for non-Euclidean spaces.

机器学习的最新进展使神经网络能够用于非欧几里德空间。.

These facilitate the learning of surface data, yet pooling strategies often remain constrained to a single fixed-graph.

这些有助于表面数据的学习，但池策略通常仅限于一个固定的图。.

This paper proposes a new learnable graph pooling method for processing multiple surface-valued data to output subject-based information.

本文提出了一种新的可学习图池方法，用于处理多个表面值数据，输出基于主题的信息。.

The proposed method innovates by learning an intrinsic aggregation of graph nodes based on graph spectral embedding.

该方法的创新之处在于基于图谱嵌入学习图节点的内在聚合。.

We illustrate the advantages of our approach with in-depth experiments on two large-scale benchmark datasets.

我们通过在两个大型基准数据集上的深入实验来说明我们方法的优势。.

The ablation study in the paper illustrates the impact of various factors affecting our learnable pooling method.

本文中的消融研究阐明了影响我们可学习池方法的各种因素的影响。.

The flexibility of the pooling strategy is evaluated on four different prediction tasks, namely, subject-sex classification, regression of cortical region sizes, classification of Alzheimer’s disease stages, and brain age regression.

在四种不同的预测任务上评估了合并策略的灵活性，即受试者性别分类、皮质区大小回归、阿尔茨海默病分期分类和脑年龄回归。.

Our experiments demonstrate the superiority of our learnable pooling approach compared to other pooling techniques for graph convolutional networks, with results improving the state-of-the-art in brain surface analysis…

我们的实验表明，与其他用于图卷积网络的池技术相比，我们的可学习池方法具有优越性，其结果改进了脑表面分析的最新技术。。.

作者(Authors)

[‘Karthik Gopinath’, ‘Christian Desrosiers’, ‘Herve Lombaert’]