NII Methods: Neural network stability

A recent publication by Mujica-Parodi et al. introduced a novel fMRI-derived metric referred to as "neural network stability" as a potential biomarker of brain aging. We will first discuss some of the pertinent findings from this publication, then walk through the steps for calculating neural network stability and explore some of the nuances in its interpretation. We will provide a working example of how this method might be utilized and expanded upon via a preliminary analysis of network stability and executive function in the setting of Parkinson's disease. This example will help illustrate some crucial caveats when attempting to conceptualize the mechanistic underpinnings of this novel biomarker. We will conclude by touching on a potential role for metabolism and directions for future research.