Areas of Expertise
A brain–computer interface (BCI) is a device that provides a direct communication channel between one’s brain and an external system. By replacing the act of reading or typing with a direct link between one’s brain and a computer, BCIs promise to significantly decrease the amount of time required to perform various information-processing tasks. At APL, we are developing practical technologies and applications to achieve this vision, largely through noninvasive neural interfaces such as electroencephalography (EEG), near-infrared spectroscopy (NIRS), functional magnetic resonance imaging (fMRI), and transcranial direct-current stimulation (tDCS).
Neuromimetic computing aims to replicate some functions of the human brain in silico through hardware and/or software implementations. These efforts are motivated by the fact that humans outperform computers on myriad analysis and inference tasks, especially those that require interpretation of complex sensory data. Therefore, if a computer can perform operations sufficiently similar to those performed in the brain, it is possible that the resulting system could have the best of both worlds: human-like cognition at machine-like speeds. At APL, we are striving to achieve this goal by developing a number of neuro-inspired algorithms and neuromorphic computer hardware systems, primarily for the purposes of automated image interpretation.
Neuroprosthetic devices interface directly with the central or peripheral nervous system to replace a motor, sensory, or cognitive deficit caused by injury or disease. For example, whereas a conventional hearing aid amplifies sounds and presents them to the intact ear, a neuroprosthetic cochlear implant converts sounds into electrical pulses that it delivers directly to the auditory nerve. At APL, we have been working on a neuroprosthetic upper-limb system that receives its commands from, and sends sensory feedback to, the brain, peripheral nerves, and/or cutaneous sites. We are currently evaluating this device and related technologies for use with upper-limb amputees, individuals with high spinal cord injuries, and other people with limited upper-limb mobility.
It is generally accepted that cognition emerges from the activity of distributed neural networks in the brain and that individual differences in performance on cognitive tasks arise from individual differences in the function and/or structure of the brain. At APL, we use neuroimaging tools (including EEG, magnetoencephalography [MEG], NIRS, and fMRI) to identify patterns of neural activity or connectivity underlying performance of various mission-relevant tasks. We then use this information to enhance our understanding of expert versus novice tradecraft, to design novel algorithms, to optimize training, and to inform selection, among other applications.