Pipeline comparisons of convolutional neural networks for structural connectomes : predicting sex across 3,152 participants

Tools

Yeung, Hon Wah and Luz, Saturnino and Cox, Simon R. and Buchanan, Colin R. and Whalley, Heather C. and Smith, Keith M.; (2020) Pipeline comparisons of convolutional neural networks for structural connectomes : predicting sex across 3,152 participants. In: 2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC). Proceedings of the Annual International Conference of the IEEE Engineering in Medicine and Biology Society, EMBS . IEEE, CAN, pp. 1692-1695. ISBN 9781728119908 (https://doi.org/10.1109/EMBC44109.2020.9175596)

[thumbnail of Yeung-etal-EMBC-2020-Pipeline-comparisons-of-convolutional-neural-networks-for-structural-connectomes]
Preview
 Text. Filename: Yeung-etal-EMBC-2020-Pipeline-comparisons-of-convolutional-neural-networks-for-structural-connectomes.pdf
Accepted Author Manuscript
License: Strathprints license 1.0
Download (615kB)| Preview

Abstract

With several initiatives well underway towards amassing large and high-quality population-based neuroimaging datasets, deep learning is set to push the boundaries of what is possible in classification and prediction in neuroimaging studies. This includes those that derive increasingly popular structural connectomes, which map out the connections (and their relative strengths) between brain regions. Here, we test different Convolutional Neural Network (CNN) models in a benchmark sex prediction task in a large sample of N=3,152 structural connectomes acquired from the UK Biobank, and compare results across different connectome processing choices. The best results (76.5% test accuracy) were achieved using Fractional Anisotropy (FA) weighted connectomes, without sparsification, and with a simple weight normalisation through division by the maximum FA value. We also confirm that for structural connectomes, a Graph CNN approach, the recently proposed BrainNetCNN, outperforms an image-based CNN.

ORCID iDs

Yeung, Hon Wah, Luz, Saturnino, Cox, Simon R., Buchanan, Colin R., Whalley, Heather C. and Smith, Keith M. ORCID logoORCID: https://orcid.org/0000-0002-4615-9020;
  • Item type:Book Section
    ID code:87316
    Dates:
    Date
    Event
    27 August 2020
    Published
    10 April 2020
    Accepted
    Notes:© 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
    Subjects:Science > Mathematics > Electronic computers. Computer science > Other topics, A-Z
    Department:Faculty of Science > Computer and Information Sciences
    Depositing user: Pure Administrator
    Date deposited:14 Nov 2023 14:13
    Last modified:13 Nov 2024 21:45
    Related URLs:
    URI:https://strathprints.strath.ac.uk/id/eprint/87316
CORE (COnnecting REpositories)