# Deep learning and whole-brain networks for biomarker discovery: modeling the dynamics of brain fluctuations in resting-state and cognitive tasks

## TL;DR

This paper explores how brain network models can be used to identify biomarkers for different brain states, such as rest and cognitive tasks, using deep learning and bifurcation parameters.

## Contribution

The study introduces bifurcation parameters from whole-brain network models as potential biomarkers for distinguishing brain states.

## Key findings

- Bifurcation parameter distributions significantly differ across task and resting-state conditions.
- A machine learning model classified predicted bifurcation values into eight cohorts with 62.63% accuracy.
- Task-based brain states exhibit higher bifurcation values compared to resting states.

## Abstract

Brain network models offer insights into brain dynamics, but the utility of model-derived bifurcation parameters as biomarkers remains underexplored. This study evaluates bifurcation parameters from a whole-brain network model as biomarkers for distinguishing brain states associated with resting-state and task-based cognitive conditions. Synthetic BOLD signals were generated using a supercritical Hopf brain network model to train deep learning models for bifurcation parameter prediction. Inference was performed on Human Connectome Project data, including both resting-state and task-based conditions. Statistical analyses assessed the separability of brain states based on bifurcation parameter distributions. Bifurcation parameter distributions differed significantly across task and resting-state conditions (\documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$p < 0.0001$$\end{document} for all but two comparisons). Task-based brain states exhibited higher bifurcation values compared to rest. At the individual level, a machine learning model was able to classify the predicted bifurcation values into eight cohorts with 62.63% accuracy (well above the 12.50% chance level). Bifurcation parameters effectively differentiate cognitive and resting states, warranting further investigation as biomarkers for brain state characterization and neurological disorder assessment.