Back

Fractal dimension distributions of resting-state EEG improve detection of dementia and Alzheimer's disease compared to traditional fractal analysis

Yoder, K. J.; Brookshire, G.; Glatt, R.; Merrill, D. A.; Gerrol, S.; Quirk, C.; Lucero, C.

2024-06-11 neurology
10.1101/2024.06.10.24308727 medRxiv
Show abstract

Across many resting-state electroencephalography (EEG) studies, dementia is associated with changes to the power spectrum and fractal dimension. Here we describe a novel method to examine changes in fractal dimension over time and within frequency bands. This method, which we call Fractal Dimension Distributions (FDD), combines spectral and complexity information. In this study, we illustrate this new method by applying it to resting-state EEG data recorded from patients with subjective cognitive impairment (SCI) or dementia. We compared the performance of FDD with the performance of standard fractal dimension metrics (Higuchi and Katz FD). FDD revealed larger group differences detectable at greater numbers of EEG recording sites. Moreover, linear models using FDD features had lower AIC and higher R2 than models using standard full time-course measures of fractal dimension. FDD metrics also outperformed the full time-course metrics when comparing SCI with a subset of dementia patients diagnosed with Alzheimers disease (AD). FDD offers unique information beyond traditional full time-course fractal analyses and may help to identify AD dementia and non-AD dementia.

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.