Abstract
Cognitive decline can be observed due to a myriad of causes. Clinicians would benefit from a noninvasive quantitative tool to screen and monitor brain function based on direct measures of neural features. In this study, we used neuroimaging data from magnetoencephalography (with a whole-head Elekta Neuromag 306 sensor system) to derive a set of features that strongly correlate with brain function. We propose that simple signal characteristics related to peak variability, timing, and abundance can be used by clinicians as a screening tool to investigate cognitive function in at-risk individuals. Using a minimalistic set of features, we were able to perfectly distinguish between participants with normative and nonnormative brain function, and we were also able to successfully predict participants’ Mini-Mental Test score (r=0.99; P<.001; mean absolute error=0.413). This set of features can be easily visualized in an analog fashion, providing clinicians with several graded measurements (in comparison to a single binary diagnostic tool) that can be used for screening and monitoring cognitive decline.
Subject
Health Informatics,Medicine (miscellaneous)