Identify Uncorrelated Channels

The uncorrelated_channels step identifies and marks EEG channels as uncorrelated if their per-epoch correlation coefficient with neighboring channels (computed across samples) is consistently a low outlier relative to the corresponding coefficients of other channels.

This step operates in two stages:

  1. For each epoch, channels whose correlation coefficient with neighboring channels (across samples) is identified as a lower outlier are flagged.

  2. a channel is marked as uncorrelated if it is flagged in more more than flag_crit proprortion of epochs.

Examples

noisy_epochs:
    flag_crit: 0.2
    outlier_method: quantile
    outliers_kwargs:
        k: 6
        lower: 0.25
        upper: 0.75

Parameters

  • flag_critfloat, default=0.2

    Proportion of epochs in which this channel must be flagged during for the channel to be flagged.

  • outlier_methodstr, default=’quantile’

    Method used to identify outlier epochs. Options are 'quantile', 'trimmed', or 'fixed'.

    • 'quantile'

      Identifies outliers using asymmetric, quantile-based thresholds. For a given epoch, channels are flagged if their standard deviation lies outside

      μ - k * (μ - q_low)

      where μ is the mean standard deviation across channels, q_low is the lower quantile specified by outliers_kwargs['lower'], and k is a scaling factor.

      This method is robust to skewed distributions and is suitable when channel variability is not symmetrically distributed.

    • 'trimmed'

      Identifies outliers using robust statistics by computing the trimmed mean and trimmed standard deviation across channels. For a given epoch, channels with standard deviation values outside

      trimmed_mean - k * trimmed_std

      are flagged.

      This method reduces the influence of extreme artifacts while preserving sensitivity to moderately noisy channels.

    • 'fixed'

      Identifies outliers using fixed, absolute thresholds. For a given epoch, channels with standard deviation values outside the thresholds (in volts) specified by outliers_kwargs['lower'] are flagged.

  • outliers_kwargsdict

    Additional keyword arguments specific to the chosen outlier detection method.

    • For 'quantile':
      • kfloat, default=6

        Scaling factor for the quantile-based thresholds.

      • lowerfloat, default=0.25

        Lower quantile.

      • upperfloat, default=0.75

        Upper quantile.

    • For 'trimmed':
      • kfloat, default=3

        Scaling factor for trimmed standard deviation thresholds.

Notes

The trimmed and fixed methods are not robustly tested in the current codebase. Please use with caution and validate results carefully.