# How to Improve the Reliability of Aperiodic Parameter Estimates in M/EEG: A Method Comparison

## TL;DR

This paper compares methods for estimating aperiodic brain activity in M/EEG data and proposes a more reliable approach to improve results.

## Contribution

A theory-driven censored regression method is proposed to improve the reliability of aperiodic parameter estimates.

## Key findings

- Allowing more peaks in fooof reduces the reliability of aperiodic parameter estimates.
- Censored regression provides more robust and reliable estimates compared to fooof.
- The method avoids overfitting and improves resistance to outliers.

## Abstract

Interest in broadband aperiodic brain activity (1/f phenomenon) has increased exponentially over recent years, partly fueled by the development of tools to parameterize it (i.e., estimate its offset/intercept and exponent/slope) using the M/EEG power spectrum. Broadband aperiodic activity needs to be separated from narrowband periodic activity before its parameters are computed. A popular method, the fooof toolbox, is based on the data‐driven detection of narrowband‐periodic peaks, whose maximum number is set by the user. While increasing analytic flexibility, variability in the number of detected peaks may increase sensitivity to noise and reduce the reliability of aperiodic parameter estimates and the power of analytic pipelines. Here, we present an investigation of the effects of analytic choices (e.g., number of peaks, spectral estimation method) on metrics indicating the adequacy of spectral parametrization. These include the internal consistency (odd‐even reliability) of aperiodic estimates, the number of outliers generated, and their ability to detect effects. Across two different data sets (resting state and task‐based), we found a decrease in the reliability of intercept and slope estimates as more peaks were allowed to be extracted. To ameliorate this problem, we propose a theory‐driven modification of fooof labeled censored regression, whereby a theory‐driven range of frequencies expected to contain periodic activity is removed from all spectra, and the remaining power values are regressed on the remaining frequencies to obtain parameter estimates. This method shows more reliable and robust estimates compared to fooof, while avoiding overfitting.

Interest in broadband aperiodic, non‐oscillatory brain activity (1/f) has increased exponentially over recent years, partly fueled by the development of tools to isolate 1/f from periodic activity and parametrize it, including the fooof toolbox. We present analyses designed to assess and maximize the reliability, statistical power, and resistance to outliers of different parametrization methods. This includes fooof with different number of peaks and a robust variant we have developed (censored regression). Our findings provide suggestions for best practices for estimating 1/f parameters.