Neuroimaging and the Brain-Computer Interface


The field of neurofunctional imaging (or neuroimaging) has evolved significantly over the past two decades. What started out as a means to confirm information derived ­from subjects suffering brain damage, has emerged as a tool for a number of possible ­security-related applications. Several neuroimaging techniques, such as positron emission tomography (PET), magnetic resonance imaging (MRI) and electro­encephalography (EEG) have been used to monitor or measure brain function.

Neuroimaging applications that may be of particular interest to the defence and national security sectors have been identified in the “Expert Assessment of Neuroimaging and Brain-Computer Interface” report by Dr. Lizann Bolinger and colleagues (October 2005):

  • brain-machine interfaces for ­instrumentation control;
  • two-way brain computer interfaces;
  • interrogation and lie detection;
  • determining intent to do harm;
  • hand-held anatomical imaging or ­functional imaging devices; and
  • remote imaging and detection.

While the overall goal of forensic imaging is to unravel the biological basis of violence and psychopathy, neuroimaging techniques are also being investigated as tools for lie detection and interrogation, and may someday provide personality assessment. The lack of ability to infer function (such as mood, thought, intention and memory) from brain structure has sparked development in the neuro­functional imaging field.

In a broad sense, functional imaging encompasses a wide variety of measurements (physiological and metabolic) such as perfusion, blood flow, metabolism, drug receptor distribution, to name just a few. However, the most prominent association people have with the term neurofunctional imaging is neural activity, or how the brain thinks.

Brain-Machine interfaces have been the primary application of functional ­neuroimaging (other than knowledge-gathering and medical diagnostics), and several companies now sell multi-electrode devices. The Brain-Computer interface is a more sophisticated approach, however, as it implies two-way communication: the person controls the computer and the computer provides responses, by methods other than visual feedback, on the computer screen. Initial attempts at this have been developed for surgical procedures, providing feedback to the surgeon through tactile sensory input. Work has not been done using direct electrical signals to the brain. Scientists are, however, working to grow brain cells on computer chips, which might act as an interface into the brain, but this approach is outside the scope of a neuroimaging assessment.

Techniques are continuously being improved and refined for medical applications and this trend is likely to continue, and these improvements will quickly translate to new applications.

The true limit in applying neurofunctional imaging to problems like lie detection, brain-computer interfaces and such, is our lack of understanding of how the physical and physiological measurements that we can make on the brain relate to those aspects of the mind that we want to detect. Cognition and the brain processes that relate to it are not well understood. Thus, for neurofunctional imaging to be used in these cases, more work needs to be performed, both on the methods used to cause brain activation (paradigms), and processing and interpreting the resulting data. This will require a deeper understanding of the mind than is currently available, but would provide the biggest technological leap and would greatly accelerate the timelines of all such applications for security interests.

Neuro­cognitive function is an area where advances could significantly affect the timelines. here too, a better understanding of how the brain works (What causes particular memories to occur in association with specific stimuli? How do biochemical processes interact with thought?) leads directly to better interpretations of neurofunctional imaging results and thus better application of these methods to a host of applications. Devel­op­ing tools to allow sophisticated tasks to be performed and monitored during neurofunctional imaging experiments will go a long way to aid in developing this understanding.

Timelines associated with the development and deployment of many of these applications are generally long (15-25 years), however, it is important to realize that the ability to make measurements in neurofunctional imaging has greatly outstripped our ability to understand what the data means. The cost of medical care is pushing the development of faster and cheaper instrumentation, and therefore, the development of neuroimaging instrumentation is heavily invested by the ­corporate and government sectors.

On the other hand, development is slow in specialty applications such as hand-held detectors, implantable devices, and the application of non-medical imaging techniques to biological systems.

Defence R&D Canada (DRDC) has strong interests in the physiological, psychological and sociological aspects of human performance and capabilities, and has some laboratory capabilities to help support these interests. DRDC is also monitoring developments at institutions, such as NRC’s Institute of Biodiagnostics, research laboratories in the academic and private sectors, as well as the US Defense Advanced Research Projects Agency (DARPA).

Improving measurement capabilities involved in neurofunctional imaging and increasing the understanding of what those images mean, in terms of the mind’s function and intent, will be of great benefit to future security applications. DRDC is ­supporting the challenging areas of understanding the human mind, and how it is related to human intent and performance.

Dr. Harold Stocker is a Defence Scientist, Science and Technology Policy, with DRCD.

Dr. Ingar Moen is a Director, S&T policy, with Defence R&D Canada.
© FrontLine Security 2006