Objective: Brain–computer interfaces (BCIs) could potentially be used to interact with pathological brain signals to intervene and ameliorate their effects in disease states. Here, we provide proof-of-principle of this approach by using a BCI to interpret pathological brain activity in patients with advanced Parkinson disease (PD) and to use this feedback to control when therapeutic deep brain stimulation (DBS) is delivered. Our goal was to demonstrate that by personalizing and optimizing stimulation in real time, we could improve on both the efficacy and efficiency of conventional continuous DBS.Methods: We tested BCI-controlled adaptive DBS (aDBS) of the subthalamic nucleus in 8 PD patients. Feedback was provided by processing of the local field potentials recorded directly from the stimulation electrodes. The results were compared to no stimulation, conventional continuous stimulation (cDBS), and random intermittent stimulation. Both unblinded and blinded clinical assessments of motor effect were performed using the Unified Parkinson's Disease Rating Scale.Results: Motor scores improved by 66% (unblinded) and 50% (blinded) during aDBS, which were 29% (p = 0.03) and 27% (p = 0.005) better than cDBS, respectively. These improvements were achieved with a 56% reduction in stimulation time compared to cDBS, and a corresponding reduction in energy requirements (p < 0.001). aDBS was also more effective than no stimulation and random intermittent stimulation.Interpretation BCI-controlled DBS is tractable and can be more efficient and efficacious than conventional continuous neuromodulation for PD. Ann Neurol 2013;74:449–457

See Moll and Engel (doi:) for a scientific commentary on this article.During pathologies like tremor, neural populations become locked into temporal configurations that reinforce neural synchrony to the point that motor function is compromised. Cagnan et al. present a novel stimulation approach to selectively control neural synchrony, and demonstrate that deep brain stimulation can be precisely timed to disrupt disease pathophysiology.

Microelectrode arrays (MEAs) are designed to monitor and/or stimulate extracellularly neuronal activity. However, the biomechanical and structural mismatch between current MEAs and neural tissues remains a challenge for neural interfaces. This article describes a material strategy to prepare neural electrodes with improved mechanical compliance that relies on thin metal film electrodes embedded in polymeric substrates. The electrode impedance of micro-electrodes on polymer is comparable to that of MEA on glass substrates. Furthermore, MEAs on plastic can be flexed and rolled offering improved structural interface with brain and nerves in vivo.MEAs on elastomer can be stretched reversibly and provide in vitro unique platforms to simultaneously investigate the electrophysiological of neural cells and tissues to mechanical stimulation. Adding mechanical compliance to MEAs is a promising vehicle for robust and reliable neural interfaces.

No abstract

Optimal decision-making requires balancing fast but error-prone and more accurate but slower decisions through adjustments of decision thresholds. Here, we demonstrate two distinct correlates of such speed-accuracy adjustments by recording subthalamic nucleus (STN) activity and electroencephalography in 11 Parkinson’s disease patients during a perceptual decision-making task; STN low-frequency oscillatory (LFO) activity (2–8 Hz), coupled to activity at prefrontal electrode Fz, and STN beta activity (13–30 Hz) coupled to electrodes C3/C4 close to motor cortex. These two correlates differed not only in their cortical topography and spectral characteristics but also in the relative timing of recruitment and in their precise relationship with decision thresholds. Increases of STN LFO power preceding the response predicted increased thresholds only after accuracy instructions, while cue-induced reductions of STN beta power decreased thresholds irrespective of instructions. These findings indicate that distinct neural mechanisms determine whether a decision will be made in haste or with caution.DOI: http://dx.doi.org/10.7554/eLife.21481.001

Neural interfaces are implanted devices that couple the nervous system to electronic circuitry. They are intended for long term use to control assistive technologies such as muscle stimulators or prosthetics that compensate for loss of function due to injury. Here we present a novel design of interface for peripheral nerves. Recording from axons is complicated by the small size of extracellular potentials and the concentration of current flow at nodes of Ranvier. Confining axons to microchannels of ~100 µm diameter produces amplified potentials that are independent of node position. After implantation of microchannel arrays into rat sciatic nerve, axons regenerated through the channels forming 'mini-fascicles', each typically containing ~100 myelinated fibres and one or more blood vessels. Regenerated motor axons reconnected to distal muscles, as demonstrated by the recovery of an electromyogram and partial prevention of muscle atrophy. Efferent motor potentials and afferent signals evoked by muscle stretch or cutaneous stimulation were easily recorded from the mini-fascicles and were in the range of 35-170 µV. Individual motor units in distal musculature were activated from channels using stimulus currents in the microampere range. Microchannel interfaces are a potential solution for applications such as prosthetic limb control or enhancing recovery after nerve injury.

High frequency deep brain stimulation of the thalamus can help ameliorate severe essential tremor. Here we explore how the efficacy, efficiency and selectivity of thalamic deep brain stimulation might be improved in this condition. We started from the hypothesis that the effects of electrical stimulation on essential tremor may be phase dependent, and that, in particular, there are tremor phases at which stimuli preferentially lead to a reduction in the amplitude of tremor. The latter could be exploited to improve deep brain stimulation, particularly if tremor suppression could be reinforced by cumulative effects. Accordingly, we stimulated 10 patients with essential tremor and thalamic electrodes, while recording tremor amplitude and phase. Stimulation near the postural tremor frequency entrained tremor. Tremor amplitude was also modulated depending on the phase at which stimulation pulses were delivered in the tremor cycle. Stimuli in one half of the tremor cycle reduced median tremor amplitude by ∼10%, while those in the opposite half of the tremor cycle increased tremor amplitude by a similar amount. At optimal phase alignment tremor suppression reached 27%. Moreover, tremor amplitude showed a non-linear increase in the degree of suppression with successive stimuli; tremor suppression was increased threefold if a stimulus was preceded by four stimuli with a similar phase relationship with respect to the tremor, suggesting cumulative, possibly plastic, effects. The present results pave the way for a stimulation system that tracks tremor phase to control when deep brain stimulation pulses are delivered to treat essential tremor. This would allow treatment effects to be maximized by focussing stimulation on the optimal phase for suppression and by ensuring that this is repeated over many cycles so as to harness cumulative effects. Such a system might potentially achieve tremor control with far less power demand and greater specificity than current high frequency stimulation approaches, and may lower the risk for tolerance and rebound.

An implantable neural interface capable of reliable long-term high-resolution recording from peripheral nerves has yet to be developed. Device design is challenging because extracellular axonal signals are very small, decay rapidly with distance from the axon, and in myelinated fibres are concentrated close to nodes of Ranvier, which are around 1 mum long and spaced several hundred micrometers apart. We present a finite element model examining the electrical behavior of axons in microchannels, and demonstrate that confining axons in such channels substantially amplifies the extracellular signal. For example, housing a 10-microm myelinated axon in a 1-cm-long channel with a 1000-microm(2) cross section is predicted to generate a peak extracellular voltage of over 10 mV. Furthermore, there is little radial signal decay within the channel, and a smooth axial variation of signal amplitude along the channel, irrespective of node location. Additional benefits include a greater extracellular voltage generated by large myelinated fibres compared to small unmyelinated axons, and the reduction of gain to unity at the end of the channel which ensures that there can be no crosstalk with electrodes in other channels nearby. A microchannel architecture seems well suited to the requirements of a peripheral nerve interface.