Brain-Computer Interfaces for Communication and Rehabilitation Using Intracortical Neuronal Activity from the Prefrontal Cortex and Basal Ganglia in Humans

Abstract

Brain-computer interfaces (BCIs) can help individuals with central nervous system (CNS) deficits recover lost function by enabling new communication channels (Wolpaw et al. 2002; Hochberg et al. 2012; Collinger et al. 2013) or by inducing and guiding adaptive plasticity for rehabilitation (Daly and Sitaram 2012; Mukaino et al. 2014). BCIs for communication—Ongoing research in BCIs for communication aims to maximize the rate of information transfer from the brain to the computer. The majority of BCIs for communication are driven by brain signals that were previ-ously known to be responsive to changes in user intention (Farwell and Donchin 1988; Pfurtscheller et al. 1997; Birbaumer et al. 2000; Allison et al. 2008). The neurophysiology underlying the brain signal and its relation to the condition necessitating the BCI are secondary to the need for a brain signal that is accessible, robust, and intuitive to control. There are only a few brain signals that meet these criteria. BCIs for rehabilitation—Ongoing research in BCIs for rehabilitation aims to maximize recovery of function after CNS trauma or disease. There are at least two mechanisms that might enhance rehabilitation through continued BCI use. First, a BCI for rehabilitation might induce Hebbian-like plasticity through co-activation of motor areas in the brain and the periphery by using the detection of motor-related brain signals to trigger peripheral sensorimotor stimulation (Mrachacz-Kersting et al. 2012; Mukaino et al. 2014). Second, a BCI for rehabilitation might facilitate recovery through operant conditioning of brain signals to push the CNS toward a state that is more permissive to traditional interventions (Pichiorri et al. 2015). In either case, the neurophysiology underlying the CNS deficit should be the primary factor in the selection of the BCI signal source. However, most BCIs for rehabil-itation use the same brain signals as BCIs for communication, without much attention to the subject-specific deficits. More work is needed to identify optimal brain signals for rehabilitative BCIs. Alternative BCI signal sources—The development of real-time functional magnetic resonance imaging (rtfMRI) has made it possible to use the metabolic signals from almost any brain region to drive a BCI (Sulzer et al. 2013). rtfMRI is less practical for communication, but it shows great potential for inducing plasticity in BCIs for rehabilitation (Shibata et al. 2011; Ruiz et al. 2013). Electrocorticography (ECoG) in patients with intractable epilepsy provides access to brain signals that are unattainable with less-invasive methods. ECoG electrode placement is dictated by clinical need and as a result most ECoG BCI studies used temporal and motor cortices (Leuthardt et al. 2004; Schalk and Leuthardt 2011); we identified one study that used prefrontal cortex (PFC) (Vansteensel et al. 2010). Recently, ECoG electrodes have been implanted in motor cortex exclusively for BCI purposes (Wang et al. 2013). Finally, a few researchers are chronically implanting microelectrodes into the brains of severely disabled participants to drive BCIs with neuronal spiking in motor (Hochberg et al. 2006; Collinger et al. 2013) and parietal cortices (Aflalo et al. 2015). There are yet other brain signal sources for BCIs that are apt for investigation. For example, the clinical procedure to implant deep brain stimulation (DBS) electrodes for the treatment of mood and motor disorders exposes PFC and records from microelectrodes in the basal ganglia (BG). PFC and BG may be useful signal sources for BCIs. At the Ottawa Hospital, we surgically implant DBS electrodes into the BG and thalamus for the treatment of Parkinson's disease (PD), dystonia, and essential tremor. We take these opportunities to investigate neuronal activity in the prefrontal cortex and basal ganglia as potential signal sources in BCIs for communication and rehabilitation.