A Comparative Study of Population-Graph Construction Methods and Graph Neural Networks for Brain Age Regression

被引:0
作者
Bintsi, Kyriaki-Margarita [1 ]
Mueller, Tamara T. [2 ]
Starck, Sophie [2 ]
Baltatzis, Vasileios [1 ,3 ]
Hammers, Alexander [3 ]
Rueckert, Daniel [1 ,2 ]
机构
[1] Imperial Coll London, Dept Comp, BioMedIA, London, England
[2] Tech Univ Munich, Fac Informat, Lab AI Med & Healthcare, Munich, Germany
[3] Kings Coll London, Biomed Engn & Imaging Sci, London, England
来源
GRAPHS IN BIOMEDICAL IMAGE ANALYSIS, AND OVERLAPPED CELL ON TISSUE DATASET FOR HISTOPATHOLOGY, 5TH MICCAI WORKSHOP | 2024年 / 14373卷
基金
英国工程与自然科学研究理事会;
关键词
Brain age regression; Population graphs; Graph Neural Networks;
D O I
10.1007/978-3-031-55088-1_6
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
The difference between the chronological and biological brain age of a subject can be an important biomarker for neurodegenerative diseases, thus brain age estimation can be crucial in clinical settings. One way to incorporate multimodal information into this estimation is through population graphs, which combine various types of imaging data and capture the associations among individuals within a population. In medical imaging, population graphs have demonstrated promising results, mostly for classification tasks. In most cases, the graph structure is pre-defined and remains static during training. However, extracting population graphs is a non-trivial task and can significantly impact the performance of Graph Neural Networks (GNNs), which are sensitive to the graph structure. In this work, we highlight the importance of a meaningful graph construction and experiment with different population-graph construction methods and their effect on GNN performance on brain age estimation. We use the homophily metric and graph visualizations to gain valuable quantitative and qualitative insights on the extracted graph structures. For the experimental evaluation, we leverage the UK Biobank dataset, which offers many imaging and non-imaging phenotypes. Our results indicate that architectures highly sensitive to the graph structure, such as Graph Convolutional Network (GCN) and Graph Attention Network (GAT), struggle with low homophily graphs, while other architectures, such as GraphSage and Chebyshev, are more robust across different homophily ratios. We conclude that static graph construction approaches are potentially insufficient for the task of brain age estimation and make recommendations for alternative research directions.
引用
收藏
页码:64 / 73
页数:10
相关论文
共 30 条
[1]   Graph-Based Deep Learning for Medical Diagnosis and Analysis: Past, Present and Future [J].
Ahmedt-Aristizabal, David ;
Armin, Mohammad Ali ;
Denman, Simon ;
Fookes, Clinton ;
Petersson, Lars .
SENSORS, 2021, 21 (14)
[2]  
Alam SB, 2014, JOINT INT CONF SOFT, P683, DOI 10.1109/SCIS-ISIS.2014.7044901
[3]   Image processing and Quality Control for the first 10,000 brain imaging datasets from UK Biobank [J].
Alfaro-Almagro, Fidel ;
Jenkinson, Mark ;
Bangerter, Neal K. ;
Andersson, Jesper L. R. ;
Griffanti, Ludovica ;
Douaud, Gwenaelle ;
Sotiropoulos, Stamatios N. ;
Jbabdi, Saad ;
Hernandez-Fernandez, Moises ;
Vallee, Emmanuel ;
Vidaurre, Diego ;
Webster, Matthew ;
McCarthy, Paul ;
Rorden, Christopher ;
Daducci, Alessandro ;
Alexander, Daniel C. ;
Zhang, Hui ;
Dragonu, Iulius ;
Matthews, Paul M. ;
Miller, Karla L. ;
Smith, Stephen M. .
NEUROIMAGE, 2018, 166 :400-424
[4]  
Bi WD, 2022, Arxiv, DOI arXiv:2209.08264
[5]   Multimodal Brain Age Estimation Using Interpretable Adaptive Population-Graph Learning [J].
Bintsi, Kyriaki-Margarita ;
Baltatzis, Vasileios ;
Potamias, Rolandos Alexandros ;
Hammers, Alexander ;
Rueckert, Daniel .
MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION, MICCAI 2023, PT VIII, 2023, 14227 :195-204
[6]  
Cole JH, 2020, NEUROBIOL AGING, V92, P34, DOI 10.1016/j.neurobiolaging.2020.03.014
[7]   Predicting brain age with deep learning from raw imaging data results in a reliable and heritable biomarker [J].
Cole, James H. ;
Poudel, Rudra P. K. ;
Tsagkrasoulis, Dimosthenis ;
Caan, Matthan W. A. ;
Steves, Claire ;
Spector, Tim D. ;
Montana, Giovanni .
NEUROIMAGE, 2017, 163 :115-124
[8]  
Cosmo Luca, 2020, Medical Image Computing and Computer Assisted Intervention - MICCAI 2020. 23rd International Conference. Proceedings. Lecture Notes in Computer Science (LNCS 12262), P643, DOI 10.1007/978-3-030-59713-9_62
[9]   Prediction of MCI to AD conversion, via MRI, CSF biomarkers, and pattern classification [J].
Davatzikos, Christos ;
Bhatt, Priyanka ;
Shaw, Leslie M. ;
Batmanghelich, Kayhan N. ;
Trojanowski, John Q. .
NEUROBIOLOGY OF AGING, 2011, 32 (12) :2322.e19-2322.e27
[10]  
Defferrard M, 2016, ADV NEUR IN, V29