Construction of Multimodal Connectional Templates of Brain Cortices for Healthy and Disordered Populations

Salma Dhifallah, Islem Rekik

Research output: Contribution to conferencePaperpeer-review

8 Downloads (Pure)

Abstract

While several research methods were developed to estimate individual-based representations of brain connectional wiring (i.e., a connectome), traditionally captured using multimodal MRI data (e.g., functional and diffusion MRI), very limited works aimed to estimate brain connectional template for a population of connectomes. Estimating well-representative brain templates is a key step in data normalization and standardization for classification and group comparison studies. Recently, a pioneering framework was proposed to estimate a connectionaltemplate for a population of unimodal brain networks. However, estimating a connectional template for a population of multimodal brain connectomes lying on different manifolds is absent. Creating such a multimodal brain connectional template can leverage and integrate complementary multimodal connections from each imaging modality, which can facilitate the comprehensive investigation of how a specific disorder affects a population of patients. To fill this gap, we propose a cluster-based multiview brain connectivity fusion framework to estimate a connectional brain template for a population of multimodal brain networks. Specifically, given a population of subjects, each with multimodal networks where each modality captures a brain connectional view, we first non-linearly fuse multimodal networks into a single fused network for each subject. Then, we cluster the fused networks to identify individuals sharing similar fused connectional traits in an unsupervised way. Next, through averaging networks in each cluster, we generate a representative connectional template. Finally, we construct the final multimodal connectional template by averaging the obtained template of all clusters. We evaluated our method on both healthy and disordered populations (with autism) and spotted differences between both connectional templates. Compared to other baseline methods, our fusion strategy achieved the best results in terms of template centeredness and representativeness.
Original languageEnglish
Pages118-122
Number of pages5
Publication statusPublished - 5 Apr 2018
EventIEEE International Symposium on Biomedical Imaging - Omni Shoreham Hotel, Washington DC, United States
Duration: 4 Apr 20187 Apr 2018
http://biomedicalimaging.org/2018/

Conference

ConferenceIEEE International Symposium on Biomedical Imaging
Abbreviated titleISBI'18
Country/TerritoryUnited States
CityWashington DC
Period4/04/187/04/18
Internet address

Keywords

  • brain morphological networks
  • cortex morphology
  • connectome
  • multimodal connectional template
  • graph fusion
  • disordered brain connectivity
  • MRI

Fingerprint

Dive into the research topics of 'Construction of Multimodal Connectional Templates of Brain Cortices for Healthy and Disordered Populations'. Together they form a unique fingerprint.

Cite this