Phenotyping Superagers Using Machine Learning Algorithms on Whole Brain Connectivity Resting-State fMRI Studies

Motivation: A significant gap remains in the literature regarding whole-brain functional analyses in superagers. Goal(s): Identify the key neural networks responsible for superagers' function connectivity abilities, which may contribute to brain resilience. Approach: We extended our previous work and explored the whole brain connectivity of superagers by using a random forest machine-learning algorithm (RF-MLA). Results: We confirmed the importance of the salience and default mode networks in classifying superagers and replicated the most discriminative nodes. Exploring the whole-brain connectivity analysis, the RF-MLA determined additional nodes in sensory hubs. This is a new finding and suggests novel avenues for investigating brain resilience. Impact: Our advanced analytical techniques validate existing findings and give new insights into sensory cortices that may be important for superagers' comprehending cognitive resilience. This could be helpful to guide future targeted interventions to optimize the efficiency of specific brain regions.