All Collections
Advanced Tutorials
Making Sense of Round Analysis
Making Sense of Round Analysis

How to understand the data and graphs that summarize training performance on the Myndlift dashboard.

Team avatar
Written by Team
Updated over a week ago

Myndlift allows you to view lots of data and graphs for a session or an individual round, but what do all these numbers and plots mean?

This article will walk you through the detailed information about your client’s neurofeedback training available from the Myndlift dashboard after s/he has completed a round.

After selecting your client and the relevant session, scroll down to the bottom of the screen, find the desired round, and click the ANALYSIS button to bring up the round analysis...

Basic Round Information

In the upper-left corner of the screen is some basic information about the round, including date, time, round duration, round points and number of round alerts.

  • Round Points gives an indication of how well the client did in terms of keeping his/her brain waves in the desired ranges (more on this later).

  • Round Alerts is a count of how many times (if any) a signal check was needed during the training due to poor signal (as can happen, for example, if the headset becomes looser).

Technical Round Information

Beneath the ROUND ANALYSIS heading is a row of boxes containing technical information about the round, including the active electrode (a.k.a. training electrode), the reference electrode, the name of the training protocol, the average positive feedback rate across the round, and an estimate of the amount of noise during the round.

Averages Table

Then comes the Averages Table. Let’s walk through the content.

The FREQUENCY column lists the frequencies trained. In the example shown, the client trained using the “Relaxation Training” protocol, in which Theta-B and Hi-Beta-B frequency bands are to be inhibited, and the Alpha band is to be enhanced.

Peeking at the ‘Amplitudes and Thresholds Over Time’ graph below, note that the Theta-B band ranges from 4-8 Hz, the Hi-Beta-B band from 20-30 Hz, and the Alpha band from 8-12 Hz. In the Relaxation Training protocol, the app gives positive feedback if the client is successful in reducing Theta-B and Hi-Beta-B activity while at the same time increasing Alpha activity.

The AVERAGE AMPLITUDE (µV) column gives the average amplitude in microvolts (µV) for each of the frequency bands to be inhibited or enhanced over the course of the round (in this case, 6 minutes).

What is ‘amplitude’ and what are ‘microvolts’?

  • Electrical activity is generated in the brain and propagates to the scalp where it is recorded by our electrodes.

  • The amount of activity is quantified by amplitude measured in microvolt (µV) units; for context, one µV equals one millionth of a volt so these are very weak electrical signals.

  • Now, the brain activity of interest to us in each frequency band is an oscillatory (repetitive) wave so the amplitude we are measuring is actually the ‘height’ of the wave in a given band (in this case, Theta-B, Hi-Beta-B, or Alpha). 

Amplitude can be computed in one of three ways. 

  • The peak-to-peak method computes the amplitude in terms of height from the tip (highest amplitude value) of one peak to the trough (lowest amplitude value) of the next peak in the oscillatory wave. 

  • The root mean square (RMS) method computes the square root of the mean averaged amplitude over time, making it less susceptible to noise.

  • Finally, the spectral density method applies the Fast Fourier Transform (FFT) to decompose the raw EEG signal into frequency bands; amplitude for a particular frequency band is equivalent to the density (weight) assigned by FFT; see below for more on FFT.

As it is designed for home application, Myndlift uses the spectral density method because it is the least susceptible to noise. 

This figure illustrates how the three amplitude calculations differ for the same data: 

For a peak-to-peak amplitude of 2.0 µV, the corresponding RMS amplitude would be 0.71 µV and spectral density amplitude would be 0.57 µV.  

So in the AVERAGE AMPLITUDE (µV) column, we can see that the average amplitude over the 6-minute round was 0.98 µV for the Theta-B band, 0.51 µV for the Hi-Beta-B band, and 0.88 µV for the Alpha band.

The AVERAGE THRESHOLD (µV) column gives the average threshold for positive feedback in microvolts (µV) for each of the frequency bands to be inhibited or enhanced over the course of the round.

When you selected/defined the protocol for your client, a positive feedback rate was set based on the success goal for each trained frequency band; for details see this article. (Positive feedback rate must be between 35 and 80%.) The AVERAGE THRESHOLD (µV) column gives the average amplitude value that the client’s brain waves needed to exceed (for enhance frequency bands) or be less than (for inhibit frequency bands) over the course of the round. For more information on thresholds, see this article.

According to the table, the average threshold over the 6-minute round was 1.25 µV for the Theta-B band, 0.79 µV for the Hi-Beta-B band, and 0.71 µV for the Alpha band.

A quick look at the AVERAGE AMPLITUDE and AVERAGE THRESHOLD columns together reveals that the client was generally able to achieve the training goals, with brain wave amplitudes beneath the threshold for the inhibit bands and above the threshold for the alpha band.

Note that in addition to its inclusion in the round analysis, the dashboard also provides the ‘Averages Table’ for each session (summarized across rounds).

Amplitudes and Thresholds Over Time Graph

The Amplitudes and Thresholds Over Time graph shows the amplitudes generated by the client (thick lines) and the accompanying thresholds (thin lines) plotted over time.

  • The amplitude/threshold for each frequency band is color-coded so you can easily see how the client’s actual brain waves compared with the threshold.

  • By clicking on the toggle buttons at the top, you can hide the client’s brainwaves (“Frequency Amplitudes” slider) and/or the thresholds (“Thresholds” slider).

Consistent with the Averages Table, we can see that the client was successful in generating amplitudes below the (inhibit) thresholds for Theta-B (green traces) and Hi-Beta (yellow traces), and exceeding the (reward) alpha threshold (blue traces). Also, the step down in the thresholds after 3 minutes reflects the setting of the threshold to be used for the remainder of the session; see this article for more details.

Note that for ease of visualization, a sliding 30-second window is applied to the amplitude traces. Thus, on average, theta was below the threshold, though it may not have been below the threshold at every time point.

Success Rates Over Time Graph

Using the same color-coding as the previous graph, the Success Rates Over Time graph shows the percent positive feedback for each trained frequency band plotted over time.

We can see that the client had the greatest, most consistent success keeping his/her Hi-Beta (yellow trace) amplitude beneath the (inhibit) threshold. The client also showed consistently high success keeping Theta-B (green trace) beneath the (inhibit) threshold. Maintaining alpha above the (enhance) threshold was somewhat more difficult for the client, though there appears to be improvement over time.

  • The overall success rate (across all three trained frequency bands) is represented by the thick black line. The overall success rate averaged over time gives the average positive feedback rate shown in the box beneath the ROUND ANALYSIS heading at the top of the page.

  • Percent noise over time is represented by the red trace. Note that the average noise rate shown beneath the ROUND ANALYSIS heading at the top of the page is computed exclusively from noise lasting >2 seconds.

Scores Table

The Scores table gives various scores that are also available to the client in the Myndlift app immediately after completion of the round.

For information on scoring, please see this article

Fast Fourier Transform Analysis Graph

The Fast Fourier Transform Analysis graph plots the average amplitude of brain waves during the round (density) for each frequency (in 1 Hz steps), as recorded from the training electrode.

  • The Fast Fourier Transform (FFT) is a mathematical calculation that converts time domain data (as shown in the previous graphs) into the frequency domain (1 Hz = once per second; 2 Hz = twice per second, etc.).

  • For each frequency, the FFT applied by Myndlift averages the amplitude across 1-second epochs (chunks) of EEG data during the entire round.

Generally, amplitude decreases as frequency increases, in part because the skull attenuates faster frequencies more than slower frequencies; a deviation from this pattern may be clinically remarkable.

Did this answer your question?