The physiology of brain–computer interfaces
The physiology of brain–computer interfaces
复制标题
脑机接口的生理学
DOI：
--
复制
发表时间：
2007
期刊：
影响因子：
--
通讯作者：
N. Birbaumer
中科院分区：
文献类型：
--
作者：
L. Cohen;N. Birbaumer
本期《生理学杂志》讨论的主题是不同学科的交叉，包括神经生理学、神经成像、神经康复、大脑可塑性和工程学：脑机接口（bci）。在过去的十年中，这些领域的新兴工作导致了对一般假设的探索，即有可能将大脑活动“翻译”为机械设备或计算机应用程序的特定运动。这个在几十年前还被认为是科幻小说的一般想法，现在正在被认真考虑，并吸引了世界各地越来越多的多学科领域的科学家。控制机械臂的运动或选择字母或单词显示在计算机屏幕上的目标利用侵入性或非侵入性记录的生理信号从一个行为的大脑包含重要的工程问题。但是，定义这一领域进展的最令人着迷的挑战，可能是生理学家记录和解码大脑活动的能力，这些活动足够详细，可以将个人的意志决定转化为对机械设备或计算机接口的具体指令。这项努力的成功可能会导致未来的应用，比如利用这些生理学上的进步，让瘫痪的人控制连接在瘫痪的手臂或腿上的假肢的运动，轮椅运动控制，以及闭锁患者与他人的交流。显然，这种生理挑战的临床相关性不能被夸大。
This issue of The Journal of Physiology addresses a theme that is at the crossroads of different disciplines including neurophysiology, neuroimaging, neurorehabilitation, brain plasticity and engineering: brain–computer interfaces (BCIs). Emerging work in these fields has led in the last decade to the exploration of the general hypothesis that it is possible to ‘translate’ brain activity into specific motions of mechanical devices or computer applications. This general idea, thought of as science fiction a few decades ago, is now seriously considered and engages an increasing number of scientists in multidisciplinary fields worldwide. The goals of controlling motion of a mechanical arm or choosing letters or words to be displayed on a computer screen utilizing invasively or non-invasively recorded physiological signals from a behaving brain encompass important engineering issues. But perhaps the most fascinating challenge that will define the progress of this field will be the ability of physiologists to record and decode brain activity with enough detail to translate volitional decisions of an individual into specific instructions to mechanical devices or computer interfaces. Success in this endeavour may lead in the future to applications like the utilization of these physiological advances to allow paralysed individuals to control movements of prostheses attached to paralysed arms or legs, wheelchair movement control, and communication in locked-in patients with others. Clearly, the clinical relevance of this physiological challenge cannot be overstated.
The invited contributions of this issue address the description of the various types of neurophysiological signals utilized to control BCI devices and the research that is under way to improve the detection tools, data processing and control of BCI applications. Eberhard Fetz focuses on the discussion of the extent to which neural signals can be volitionally controlled, particularly measured by the activity of cortical neurons in for example operant conditioning paradigms using biofeedback. He also discusses the limits in the degree of accuracy of control obtained from physiological signals recorded in recent human studies (Fetz, 2007). Andy Schwartz discusses how behavioural aspects of action are represented in motor cortical activity, focusing on the extent to which kinematic parameters of movement relate to neural activity in the motor cortex (Schwartz, 2007). Fetz's and Schwartz's articles review the physiological principles that represent the foundations of human BCI applications.
John Donoghue's article describes human applications of these principles using signals recorded directly from the motor cortex with intracortical microelectrodes (Donoghue et al. 2007). The representations of body parts deafferented or deefferented by lesions like stroke or spinal cord injury persist in the motor cortex even years after injury. These signals could be used by paralysed individuals to operate a range of mechanical or computer devices. Jonathan Wolpaw's article focuses on the use of physiological signals recorded non-invasively from the brain (Wolpaw, 2007). He discusses the importance of the principles of cooperativity of multiple cortical regions to generate a particular behaviour and the involvement of adaptive plasticity on the choice of the particular physiological signals to control BCI devices. Niels Birbaumer focuses on the link between physiological and clinical applications of non-invasive brain–computer interfaces like brain communication in paralysis and motor restoration in stroke (Birbaumer & Cohen, 2007). Finally, Bruce Dobkin's article assesses the impact of BCI work performed so far in the basic and clinical domains on the field of clinical neurorehabilitation, as well as the future challenges posed by these patient populations (Dobkin, 2007).
Altogether, these papers provide a comprehensive review of the physiological principles of brain–computer interfaces, the reality, the physiological challenges and the expected future developments, as well as their clinical implications in the fields of neurorehabilitation and motor control.