Neurofeedback is an integral part of many strategies in clinical neuroengineering, such as closed-loop control and active methods to get

participants to explicitly control their own brain activity.

We consider three components of neurofeedback systems: first, methods to measure, analyse and interpret brain and behaviour data using real-time technology; second, design of algorithms used on the part of the control system to govern the effector response, which can vary from simple control to autonomous control algorithms; and third, the presence and nature of learning and adaptive changes in the brain – so called co-adaptive learning. By considering these processes together, we can understand and design optimised closed-loop interventions for a broad variety of applications.