A brain–computer interface working definition

Abstract

We read with interest the Comment that suggests a taxonomy for brain–computer interfaces (BCIs) based on application¹, and we agree that a nomenclature of different types of BCI may be useful for discussing this broad and expanding field comprised of stakeholders with different backgrounds and expertise, including groups new to BCI, such as media, investors and students. Nevertheless, we feel compelled to respond regarding the working definition of BCIs proposed in the article. The definition endorsed by the authors was that a BCI includes “any technology that records brain activity and processes it on an electronic device, or any technology that stimulates brain activity based on computations performed on an electronic device”. Although this may be accurate on a strictly semantic basis — that is, recording and stimulating technically meet the definition of an interface as both involve interactions between two systems — it diverges both from the historical definition of BCIs² and from the results obtained in a consensus survey recently performed by the BCI Society. The survey was sent to a broad range of stakeholders in the field, including scientists, engineers, physicians, therapists, regulators and industry members. Based on the results of this survey and continued discussion, 57% of 137 respondents, all recent or current members of the BCI Society, agreed upon a working definition, reflecting the potential need to revisit this important topic in the future. This working definition was carefully created by an ad hoc committee of the BCI Society, as follows: “A brain-computer interface is a system that measures brain activity and converts it in (nearly) real-time into functionally useful outputs to replace, restore, enhance, supplement, and/or improve the natural outputs of the brain, thereby changing the ongoing interactions between the brain and its external or internal environments. It may additionally modify brain activity using targeted delivery of stimuli to create functionally useful inputs to the brain”.

We believe that this definition captures more clearly the essential features of a BCI and helps to distinguish it from other technologies that interface with or record from the brain, such as clinical electroencephalography and magnetoencephalography used for diagnostic purposes, functional magnetic resonance imaging or nuclear medicine scans, that are all consistent with the recently proposed definition¹. These technologies all record brain activity and perform computations on electronic devices, yet no one would consider them BCIs. Other devices such as open-loop deep brain stimulators could also be included under the proposed BCI definition¹, as they stimulate the brain based on computations performed on an electronic device. Yet, the responses to the BCI Society survey indicate an overall preference for a definition that includes measuring and responding to brain activity, not simply stimulating the brain. While a broadly inclusive definition can help to draw media and investors’ attention, which is important to the success of the field, it might cause confusion for regulators and insurers, as well as patients and clinicians. We additionally assert the critical importance of including in the definition the rapid interaction between the brain and its environment (whether internal or external), as established in the working definition from the BCI Society but not reflected in the recently proposed definition¹.