Use of a Bayesian maximum-likelihood classifier to generate training data for brain–machine interfaces

Abstract

Brain-machine interface decoding algorithms need to be predicated on assumptions that are easily met outside of an experimental setting to enable a practical clinical device. Given present technology limitations, there is a need for decoding algorithms which a) are not dependent upon a large number of neurons for control, b) are adaptable to alternative sources of neuronal input such as local field potentials, and c) require only marginal training data for daily calibrations. Moreover, practical algorithms must recognize when the user is not intending to generate a control output and eliminate poor training data. In this study, we introduce and evaluate a Bayesian Maximum-Likelihood Estimation (bMLE) strategy to address the issues of isolating quality training data and self-paced control. Six animal subjects demonstrate that a multiple state classification task, loosely based on the standard center-out task, can be accomplished with fewer than five engaged neurons while requiring less than ten trials for algorithm training. In addition, untrained animals quickly obtained accurate device control utilizing local field potentials as well as neurons in cingulate cortex, two non-traditional neural inputs.

Reliability of directional information in unsorted spikes and local field potentials recorded in human motor cortex

“…Directional information in LFPs is reported in rat (Slutzky et al, 2011, Ludwig et al, 2011) and monkey motor cortex (Ince et al, 2010, Rickert et al, 2005, Donoghue et al., 1998, Zhuang et al, 2010a) with high decoding accuracy at low (<5Hz) and high frequencies (>63Hz), while the 16–42Hz band showed poor directional tuning (Rickert et al, 2005). In addition, our results show that including LFP frequencies above 200Hz may be beneficial (Zhuang et al, 2010a).…”

PEDOT–CNT Composite Microelectrodes for Recording and Electrostimulation Applications: Fabrication, Morphology, and Electrical Properties

“…Their excellent electronic and ionic conductivity providing for low impedance and high capacitance (Skotheim, 1998; Green et al, 2008; Inzelt, 2008; Ludwig et al, 2011), the ease of fabrication by electropolymerization (Skotheim, 1998; Inzelt, 2008), their suitability for the construction of sensors (Schuhmann, 1995; Bobacka et al, 2003; Cosnier, 2003; Bobacka, 2006; Bai and Shi, 2007; Lange et al, 2008; Ates and Sarac, 2009) as well as of devices for neuronal stimulation and recording (Richardson-Burns et al, 2007a,b; Abidian et al, 2009, 2010; Wilks et al, 2009; Egeland et al, 2010) along with excellent biostability and biocompatibility in principle suggest conducting polymers as useful electrode materials for various uses in neurobiological research.…”

Design Choices for Next-Generation Neurotechnology Can Impact Motion Artifact in Electrophysiological and Fast-Scan Cyclic Voltammetry Measurements

“…This subtraction is often known as ‘common-mode’ rejection. Electrophysiological recordings in vivo have historically used rigid multi-electrode arrays or tethered microwire systems [ 9 , 10 , 11 , 25 , 53 , 59 , 61 , 62 ], which have several advantages with respect to minimizing motion artifact. First, the rigid structures minimize electrical artifacts by resisting bending, and therefore limit artifacts described above [ 25 ].…”

“…The information recorded from these instruments can also enable state-of-the-art therapeutic or rehabilitative strategies [ 8 ]. For example, extracellular neuronal signal information recorded from implanted microelectrode arrays (i.e., single units, multiunits, or local field potentials) may be used to decode intended movements and control assistive devices [ 1 , 9 , 10 , 11 , 12 ]. Similarly, measurements of phasic changes in the extracellular concentration of dopamine taken by small electrodes placed within the brain have been proposed as a feedback signal to titrate levels of deep brain stimulation (DBS) to alleviate tremors associated with Parkinson’s Disease [ 13 , 14 , 15 ].…”