Research
Print page Print page
Switch language
The Capital Region of Denmark - a part of Copenhagen University Hospital
E-pub ahead of print

Generalizability of machine learning for classification of schizophrenia based on resting-state functional MRI data

Research output: Contribution to journalJournal articleResearchpeer-review

DOI

  1. Classification of social anhedonia using temporal and spatial network features from a social cognition fMRI task

    Research output: Contribution to journalJournal articleResearchpeer-review

  2. Transcranial direct current stimulation over the sensory-motor regions inhibits gamma synchrony

    Research output: Contribution to journalJournal articleResearchpeer-review

  3. Trait Openness and serotonin 2A receptors in healthy volunteers: A positron emission tomography study

    Research output: Contribution to journalJournal articleResearchpeer-review

  4. Subclinical depressive symptoms during late midlife and structural brain alterations: A longitudinal study of Danish men born in 1953

    Research output: Contribution to journalJournal articleResearchpeer-review

  1. Classification of social anhedonia using temporal and spatial network features from a social cognition fMRI task

    Research output: Contribution to journalJournal articleResearchpeer-review

  2. Accessibility of cortical regions to focal TES: Dependence on spatial position, safety, and practical constraints

    Research output: Contribution to journalJournal articleResearchpeer-review

  3. Electric field simulations for transcranial brain stimulation using FEM: an efficient implementation and error analysis

    Research output: Contribution to journalJournal articleResearchpeer-review

  4. No trace of phase: Corticomotor excitability is not tuned by phase of pericentral mu-rhythm

    Research output: Contribution to journalJournal articleResearchpeer-review

  • Xin-Lu Cai
  • Dong-Jie Xie
  • Kristoffer H Madsen
  • Yong-Ming Wang
  • Sophie Alida Bögemann
  • Eric F C Cheung
  • Arne Møller
  • Raymond C K Chan
View graph of relations

Machine learning has increasingly been applied to classification of schizophrenia in neuroimaging research. However, direct replication studies and studies seeking to investigate generalizability are scarce. To address these issues, we assessed within-site and between-site generalizability of a machine learning classification framework which achieved excellent performance in a previous study using two independent resting-state functional magnetic resonance imaging data sets collected from different sites and scanners. We established within-site generalizability of the classification framework in the main data set using cross-validation. Then, we trained a model in the main data set and investigated between-site generalization in the validated data set using external validation. Finally, recognizing the poor between-site generalization performance, we updated the unsupervised algorithm to investigate if transfer learning using additional unlabeled data were able to improve between-site classification performance. Cross-validation showed that the published classification procedure achieved an accuracy of 0.73 using majority voting across all selected components. External validation found a classification accuracy of 0.55 (not significant) and 0.70 (significant) using the direct and transfer learning procedures, respectively. The failure of direct generalization from one site to another demonstrates the limitation of within-site cross-validation and points toward the need to incorporate efforts to facilitate application of machine learning across multiple data sets. The improvement in performance with transfer learning highlights the importance of taking into account the properties of data when constructing predictive models across samples and sites. Our findings suggest that machine learning classification result based on a single study should be interpreted cautiously.

Original languageEnglish
JournalHuman Brain Mapping
ISSN1065-9471
DOIs
Publication statusE-pub ahead of print - 1 Oct 2019

    Research areas

  • generalizability, machine learning, reproducibility, schizophrenia spectrum disorders

ID: 58060515