Neurofeedback and Biofeedback: Uses and Configurations

Biofeedback systems may use heart rate, heart-rate variability (HRV), the beat-to-beat variation in the interval between successive heartbeats, respiration, electrodermal activity (EDA), changes in skin conductance driven mainly by sweat-gland activity, muscle activity or skin temperature. EEG neurofeedback may use spectral power, the strength of EEG activity within a defined frequency band, sensorimotor rhythms or other predefined electrophysiological features.

Feedback can be visual, auditory, game-based or presented in VR. The scientific value of the system depends on whether the target physiological variable is clearly defined and measured reliably in real time. [1]

During neurofeedback, the measured signal is processed continuously and translated into external feedback. The participant can then adjust strategy or state while the system updates the feedback. Signal quality, processing algorithms, feedback delay and learning design all influence the training loop.

The CRED-nf consensus recommends transparent reporting of controls, preregistration, signal processing, neurophysiological outcomes and behavioural outcomes in neurofeedback research. [2]

Feedback systems are selected according to the physiological variable that is meant to be learned or regulated, and according to whether the goal is research, rehabilitation, clinical training or performance.

Neurofeedback has been studied across neurological and psychiatric conditions, but the quality of evidence and effect sizes vary by population, target signal, control condition and training protocol. [3]

HRV biofeedback has been evaluated in systematic reviews and meta-analyses for stress, anxiety and physiological self-regulation. [4,5] These findings still need to be interpreted in relation to the specific protocol and comparison condition.

For research use, access to raw signals, filters, thresholds, feedback latency and data export is valuable. EEG neurofeedback also depends on electrode contact and artefact handling; HRV biofeedback depends on reliable cardiac and respiratory signals.

A system intended for research should make enough of this processing visible to reproduce the protocol and understand what the participant is actually being trained to change.

Configurations differ by signal, number of channels, feedback modality and transparency of real-time processing. Research systems generally benefit from raw-data access and configurable algorithms.