Specifying and Developing References for Live Z-Score Neurofeedback

STS Instructor Authored by Thomas Collura

Article at a glance

Practitioner perspective. This article describes a clinician's own experience, cases or methods. It was not peer reviewed or controlled, so it can be subjective. Read it as practical insight, not as evidence that a treatment works.

Area: Practitioner Perspectives
Authors: Collura TF
Journal: NeuroConnections (Spring 2014)
Year: 2014
Protocol: Live z-score training; choice of normative, individual or custom reference databases
Equipment: BrainMaster BrainAvatar; NeuroGuide; Lifespan (ANI) and BrainDX reference databases

Key findings

A z-score reference is just a set of means and standard deviations, so it need not come from a 'normal' population
Individual, task-based or goal-based references are proposed alongside normative databases
A normal person with fast, non-sinusoidal alpha can look abnormal (false excess beta) on a standard z-score map
Technical and conceptual; no treatment outcome data

An article questioning whether a database of "normal" people is always the right target for live z-score training. A z-score only says how far a measurement is from a chosen mean, in standard deviations. The author argues that the reference could just as well be a chosen group, an ideal state, or the client's own EEG in a better state. He gives the example of a healthy, high-performing person whose fast, non-sinusoidal alpha rhythm makes a standard z-score map show false excess beta. Training that person toward the norm could be irrelevant or even harmful. He also covers the effects of age, eyes-open versus task recordings, and how dynamic (moment-to-moment) and static references relate, showing close agreement between his company's filter-based measures and FFT-based ones. It is a technical and conceptual article; it reports no treatment outcomes.

Read it in NeuroConnections, Spring 2014 (ISNR archive), p. 26 →