Z-Score EEG Biofeedback: Past, Present, and Future
Human electroencephalogram (EEG) biofeedback (neurofeedback) started in the 1940s using one EEG recording channel, then four channels in the 1990s, and in 2004, expanded to 19 channels using Low Resolution Electromagnetic Tomography (LORETA) of the microampere three-dimensional current sources of the EEG. In 2004–2006 the concept of a real-time comparison of the EEG to a healthy reference database was developed and tested using surface EEG z score neurofeedback based on a statistical bell curve called real-time z scores. The real-time or live normative reference database comparison was developed to help reduce the uncertainty of what threshold to select to activate a feedback signal and to unify all EEG measures to a single value (i.e., the distance from the mean of an age-matched reference sample). In 2009 LORETA z score neurofeedback further increased specificity by targeting brain network hubs referred to as Brodmann areas. A symptom checklist program to help link symptoms to dysregulation of brain networks based on fMRI and positron emission tomography (PET) and neurology was created in 2009. The symptom checklist and National Institutes of Health–based networks linking symptoms to brain networks grew out of the human brain mapping program started in 1990 that continues today. A goal is to increase specificity of EEG biofeedback by targeting brain network hubs and connections between hubs likely linked to the patient's symptoms. Developments first introduced in 2017 provide increased resolution of three-dimensional source localization with 12,700 voxels using swLORETA with the capacity to conduct cerebellar neurofeedback and neurofeedback of subcortical brain hubs such as the thalamus, amygdala, and habenula. Future applications of swLORETA z score neurofeedback represent another example of the transfer of knowledge gained by the human brain mapping initiatives to further aid in helping people with cognition problems as well as balance problems and parkinsonism. A brief review of the past, present, and future predictions of z score neurofeedback are discussed with special emphasis on new developments that point toward a bright and enlightened future in the field of EEG biofeedback.
Difference between standard neurofeedback vs z score neurofeedback.
Top row is conventional or standard EEG biofeedback in which different units of measurement are used in an EEG analysis (e.g., μV for amplitude, theta/beta ratios, relative power 0 to 100%, coherence 0 to 1, phase in degrees or radians, etc.) and the clinician must guess at a threshold for a particular electrode location and frequency and age for when to reinforce or inhibit a give measure. The bottom row is z score biofeedback, in which different metrics are represented by a single and common metric (i.e., the metric of a z score, and the guesswork is removed because all measures are reinforced to move z scores toward z = 0, which is the approximate center of an average healthy brain state based on a reference age-matched normative database in real time). Reprinted from Handbook of Quantitative Electroencephalography and EEG Biofeedback by Thatcher, R. W., 2012, St. Petersburg, FL: Anipublishing.
Illustrates the historical development of LORETA from 1994 to 2007. The 1994 LORETA used 2,395 average MRI voxels and a spherical head model. The 2001 Key Institute sLORETA used 6,200 MRI voxels and a spherical head model. The swLORETA used 12,700 MRI voxels from a nonaveraged or a single MRI and a homogeneous lead field (like used with MEG) and a boundary element method rather than a spherical head model. Reprinted from Handbook of quantitative electroencephalography and EEG biofeedback (2nd ed.) by Thatcher, R. W., 2016, St. Petersburg, FL: Anipublishing.
Comparison of EEG source localization accuracy between sLORETA and swLORETA.
A comparison of the localization accuracy of sLORETA versus swLORETA. The x-axis is the signal-to-noise ratio and the y-axis is error measurements. Reprinted from Functional imaging based on swLORETA and phase synchronization by Palmero-Soler, 2010, available at https://www.appliedneuroscience.com/PDFs/Ernezto_Soler_2010_Functional_Imaging_based_on_swLORETA.pdf
An example of swLORETA inside of a navigational platform called the NeuroNavigator that allows one to navigate through MRI slices and the MRI volume to view current sources and functional and effective connectivity. This includes a symptom checklist and brain networks known to be linked to symptoms based on the human brain mapping program and publications listed in the National Library of Medicine (Pubmed). Left is the three-dimensional volume view that includes a semi-transparent cortex, diffusion tensor imaging and coherence between the hubs (Brodmann areas) of the dorsal attention network. Right is the two-dimensional “Connectome” of the dorsal attention network selected as one of several possible brain networks as established by human brain mapping fMRI and PET.
Examples of changes in z scores over neurofeedback sessions from different clinicians from their clinical practices from patients with different clinical problems. The y-axis shows z score values and the x-axis shows neurofeedback sessions in six different subjects provided by EEG biofeedback clinicians using surface and/or LORETYA z score neurofeedback to train patients.
Examples of reduced z score values in EEG brain maps in six different subjects in 10 sessions or less from four different clinicians, measured from their clinical practice using EEG z score neurofeedback.
Functional and effective connectivity in patient with sharp waves in right temporal lobe.
Example of functional (zero phase lag coherence, lagged coherence, and phase difference) and effective connectivity (phase-slope index) between all brain network hubs. This figure illustrates the use of electrical neuroimaging in epilepsy patients where the focal epileptic event is in the right posterior temporal regions. The network analyses allow one to evaluate the local and distant effects on different functional networks and then to evaluate changes over time as a function of treatments.
Cerebellum structural connections and swLORETA real-time functional connections.
The image on the left illustrates the anatomical connections of the human cerebellum. On the right is an example of the cerebellum nodes and connections to the sensory-motor cortex using the swLORETA NeuroNavigator. The image on the right also shows z scores of the EEG on the scalp surface as well as z scores of functional connectivity between the 13 hubs of the cerebellum, plus the red nucleus, subthalamus and thalamus and the somatosensory cortex. See Table 3 for a list of the swLORETA neurofeedback protocol options. (NeuroGuide v. 3.0.7, Applied Neuroscience, Inc., 2019).
Examples of cerebellar EEG biofeedback hubs and connections.
Examples of swLORETA source localization and functional and effective connectivity between cerebellar sources and the sensory-motor cortex.
Robert Thatcher
Joel Lubar
J. Lucas Koberda
Contributor Notes