What Is Neurofeedback?

Neurofeedback is a form of biofeedback in which real-time information about a person's brain activity is fed back to that person through visual, auditory, or tactile cues, enabling voluntary self-regulation of specific neural signals. The most common implementation uses electroencephalography (EEG) to extract oscillatory components of brain activity and present them as a game score, animation, or tone that the subject can learn to influence through mental effort. The goal is to shift neural patterns toward states associated with improved performance, symptom reduction, or psychological well-being. Neurofeedback protocols have been in clinical and research use since the 1960s, when Joseph Kamiya demonstrated that subjects could learn to increase occipital alpha power on demand, and Barry Sterman showed that training of sensorimotor rhythms reduced seizure frequency in cats and later in human patients with epilepsy.

EEG-Based Protocols and Frequency Bands

Neurofeedback training is organized around frequency bands of the EEG. Theta activity (4-8 Hz) is associated with drowsiness and inattention; beta activity (12-30 Hz) with alert, focused cognition; alpha activity (8-12 Hz) with relaxed wakefulness; and sensorimotor rhythms (12-15 Hz, also called SMR) with quiet attentiveness. Protocols for ADHD typically reward SMR or beta and inhibit theta at scalp electrodes over the sensorimotor cortex. Protocols for anxiety and post-traumatic stress often aim to increase alpha or theta power at frontal sites. A comprehensive review of neurofeedback methodology and clinical applications covering system design, electrode placement, and outcome evidence describes how these protocol choices map onto target symptom profiles.

fMRI Neurofeedback and Real-Time Brain Imaging

Beyond EEG, neurofeedback can be implemented with functional MRI (fMRI), which provides spatial resolution sufficient to target individual brain regions such as the anterior insula, the amygdala, or the anterior cingulate cortex. In real-time fMRI neurofeedback, blood-oxygen-level-dependent signal from a specified region of interest is computed on a TR-by-TR basis and displayed as a thermometer or similar gauge. Subjects learn to increase or decrease activity in that region, with applications in chronic pain, depression, and emotion regulation. The tradeoff compared with EEG neurofeedback is cost and restricted movement, but the ability to target deeper structures expands the range of treatable conditions. The IEEE Engineering in Medicine and Biology Society's overview of EEG-based brain-computer interfaces places neurofeedback within the broader architecture of BCI systems.

Clinical Evidence and Limitations

The clinical evidence base for neurofeedback varies substantially by condition. Its use in ADHD has the most accumulated trials, with multiple meta-analyses reporting moderate reductions in inattention and hyperactivity, though debates about active versus sham controls persist. Evidence for epilepsy management, anxiety, and PTSD is suggestive but thinner. A recurring methodological challenge is the difficulty of constructing an adequate sham control: a neurofeedback session that feels identical but delivers randomized or yokedback signals from a different participant rather than the subject's own brain. Research on EEG-neurofeedback cited in work on brain-computer interfaces and neurofeedback clinical applications identifies this control problem as the central obstacle to advancing neurofeedback from a promising intervention to an established one.

Applications

Neurofeedback has applications in a range of fields, including:

• Attention deficit hyperactivity disorder treatment as a nonpharmacological intervention
• Epilepsy management in patients with medication-refractory seizures
• Post-traumatic stress disorder and anxiety treatment
• Peak performance training in athletes, musicians, and surgeons
• Motor rehabilitation following stroke or spinal cord injury
• Sleep disorder management through slow-wave sleep enhancement protocols