EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Flexible parametrization of graph‐theoretical features from individual‐specific networks for prediction

Mariella Gregorich, Sean L Simpson, Georg Heinze iD

Statistics in Medicine · 2024

DOI: 10.1002/sim.10091 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Statistics in Medicine, 2024. Verfasst von Mariella Gregorich und 2 weiteren Autor:innen. Beteiligte Einrichtung: Medical University of Vienna, Center for Medical Data Science, Institute of Clinical Biometrics, Vienna, Austria. Hauptthemen laut MEDLINE: Magnetic Resonance Imaging, Computer Simulation. Die Arbeit nennt 33 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Österreichischen Akademie der Wissenschaften.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Statistical techniques are needed to analyze data structures with complex dependencies such that clinically useful information can be extracted. Individual‐specific networks, which capture dependencies in complex biological systems, are often summarized by graph‐theoretical features. These features, which lend themselves to outcome modeling, can be subject to high variability due to arbitrary decisions in network inference and noise. Correlation‐based adjacency matrices often need to be sparsified before meaningful graph‐theoretical features can be extracted, requiring the data analysts to determine an optimal threshold. To address this issue, we propose to incorporate a flexible weighting function over the full range of possible thresholds to capture the variability of graph‐theoretical features over the threshold domain. The potential of this approach, which extends concepts from functional data analysis to a graph‐theoretical setting, is explored in a plasmode simulation study using real functional magnetic resonance imaging (fMRI) data from the Autism Brain Imaging Data Exchange (ABIDE) Preprocessed initiative. The simulations show that our modeling approach yields accurate estimates of the functional form of the weight function, improves inference efficiency, and achieves a comparable or reduced root mean square prediction error compared to competitor modeling approaches. This assertion holds true in settings where both complex functional forms underlie the outcome‐generating process and a universal threshold value is employed. We demonstrate the practical utility of our approach by using resting‐state fMRI data to predict biological age in children. Our study establishes the flexible modeling approach as a statistically principled, serious competitor to ad‐hoc methods with superior performance.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Magnetic Resonance ImagingComputer SimulationHumansChildBrainModels, StatisticalAutistic Disorder

Stichwörter der Autor:innen

complex systemsfMRInetworkspredictionsplines

Autor:innen und Institutionen

  • Medical University of Vienna, Center for Medical Data Science, Institute of Clinical Biometrics, Vienna, Austria.
  • Department of Biostatistics and Data Science, Wake Forest University School of Medicine, Winston-Salem, North Carolina, USA.

ORCID-Profile: Georg Heinze

Förderung

  • Österreichischen Akademie der Wissenschaften

Zitieren

APA

Gregorich, M., Simpson, S. L., & Heinze, G. (2024). Flexible parametrization of graph‐theoretical features from individual‐specific networks for prediction. Statistics in Medicine. https://doi.org/10.1002/sim.10091

Vancouver

Gregorich M, Simpson SL, Heinze G. Flexible parametrization of graph‐theoretical features from individual‐specific networks for prediction. Statistics in Medicine. 2024. doi:10.1002/sim.10091

BibTeX
@article{Gregorich2024,
  author = {Gregorich, Mariella and Simpson, Sean L and Heinze, Georg},
  title = {Flexible parametrization of graph‐theoretical features from individual‐specific networks for prediction},
  journal = {Statistics in Medicine},
  year = {2024},
  doi = {10.1002/sim.10091}
}
Kostenlos starten →