EUVIMEDObjevte EUVIMED
Open access · CC BY 4.0

Frontotemporal dementia subtyping using machine learning, multivariate statistics and neuroimaging

Amelie Metz iD, Yashar Zeighami, Simon Ducharme iD, Sylvia Villeneuve, Mahsa Dadar

Brain Communications · 2024-12-24 · 7 (1)

DOI: 10.1093/braincomms/fcaf065 ↗

At a glance

Journal article in Brain Communications, published online on 2025-02-11. Written by Amelie Metz and 4 co-authors. Institution: Douglas Research Center, Montreal, Canada H4H 1R3. Cited 16 times so far (Crossref); 70 references. The full text is freely available (open access, CC BY 4.0). Funded by: Healthy Brains for Healthy Lives, Fonds de Recherche du Québec - Santé, NSERC.

AbstraktSubjectsBibliografické údajeCiteRelated studies

Abstract (original language)

Frontotemporal dementia (FTD) is a prevalent form of early-onset dementia characterized by progressive neurodegeneration and encompasses a group of heterogeneous disorders. Due to overlapping symptoms, diagnosis of FTD and its subtypes still poses a challenge. Magnetic resonance imaging (MRI) is commonly used to support the diagnosis of FTD. Using machine learning and multivariate statistics, we tested whether brain atrophy patterns are associated with severity of cognitive impairment, whether this relationship differs between the phenotypic subtypes and whether we could use these brain patterns to classify patients according to their FTD variant. A total of 136 patients (70 behavioural variant FTD, 36 semantic variant primary progressive aphasia and 30 non-fluent variant primary progressive aphasia) from the frontotemporal lobar degeneration neuroimaging initiative (FTLDNI) database underwent brain MRI and clinical and neuropsychological examination. Deformation-based morphometry, which offers increased sensitivity to subtle local differences in structural image contrasts, was used to estimate regional cortical and subcortical atrophy. Atlas-based associations between atrophy values and performance across different cognitive tests were assessed using partial least squares. We then applied linear regression models to discern the group differences regarding the relationship between atrophy and cognitive decline in the three FTD phenotypes. Lastly, we assessed whether the combination of atrophy and cognition patterns in the latent variables identified in the partial least squares analysis could be used as features in a machine learning model to predict FTD subtypes in patients. Results revealed four significant latent variables that combined accounted for 86% of the shared covariance between cognitive and brain atrophy measures. Partial least squares-based atrophy and cognitive patterns predicted the FTD phenotypes with a cross-validated accuracy of 89.12%, with high specificity (91.46–97.15%) and sensitivity (84.19–93.56%). When using only MRI measures and two behavioural tests in the partial least squares and classification algorithms, ensuring clinical feasibility, our model was equally precise in the same participant sample (87.18%, specificity 76.14–92.00%, sensitivity 86.93–98.26%). Here, including only atrophy or behaviour patterns in the analysis led to prediction accuracies of 69.76% and 76.54%, respectively, highlighting the increased value of combining MRI and clinical measures in subtype classification. We demonstrate that the combination of brain atrophy and clinical characteristics and multivariate statistical methods can serve as a biomarker for disease phenotyping in FTD, whereby the inclusion of deformation-based morphometry measures adds to the classification accuracy in the absence of extensive clinical testing.

Subjects

Author keywords

classificationfrontotemporal dementiamachine learningmagnetic resonance imagingneurodegeneration

Authors and institutions

  • Douglas Research Center, Montreal, Canada H4H 1R3.
  • Department of Psychiatry, McGill University, Montreal, Canada H3A 1A1.
  • McConnell Brain Imaging Centre, Montreal Neurological Institute, McGill University, Montreal, Canada H3A 2B4.

ORCID profiles: Amelie Metz, Simon Ducharme

Funding

  • Healthy Brains for Healthy Lives
  • Fonds de Recherche du Québec - Santé
  • NSERC
  • Biogen
  • Novo Nordisk
  • Janssen
  • Alnylam
  • Wave Life Sciences

Cite

APA

Metz, A., Zeighami, Y., Ducharme, S., Villeneuve, S., & Dadar, M. (2024). Frontotemporal dementia subtyping using machine learning, multivariate statistics and neuroimaging. Brain Communications, 7(1). https://doi.org/10.1093/braincomms/fcaf065

Vancouver

Metz A, Zeighami Y, Ducharme S, Villeneuve S, Dadar M. Frontotemporal dementia subtyping using machine learning, multivariate statistics and neuroimaging. Brain Communications. 2024;7(1). doi:10.1093/braincomms/fcaf065

BibTeX
@article{Metz2024,
  author = {Metz, Amelie and Zeighami, Yashar and Ducharme, Simon and Villeneuve, Sylvia and Dadar, Mahsa},
  title = {Frontotemporal dementia subtyping using machine learning, multivariate statistics and neuroimaging},
  journal = {Brain Communications},
  year = {2024},
  volume = {7},
  number = {1},
  doi = {10.1093/braincomms/fcaf065}
}
Start for free →