Flexible, scalable, high channel count stereo-electrode for recording in the human brain

Abstract

Over the past decade, stereotactically placed electrodes have become the gold standard for deep brain recording and stimulation for a wide variety of neurological and psychiatric diseases. Current electrodes, however, are limited in their spatial resolution and ability to record from small populations of neurons, let alone individual neurons. Here, we report on an innovative, customizable, monolithically integrated human-grade flexible depth electrode capable of recording from up to 128 channels and able to record at a depth of 10 cm in brain tissue. This thin, stylet-guided depth electrode is capable of recording local field potentials and single unit neuronal activity (action potentials), validated across species. This device represents an advance in manufacturing and design approaches which extends the capabilities of a mainstay technology in clinical neurology.

Fig. 1. µSEEG electrode arrays.

a Photograph of a single glass substrate plate with four µSEEG electrodes. b Scanning electron microscope (SEM) image of a single PtNR contact; Inset is a magnified image showing the PtNRs. c Structural composition of the µSEEG array. d Photograph showing the ‘neck’ of the array where the U-shape pattern is flipped to provide metal trace extension and circular holes are present to stabilize the inserted stainless-steel stylet. e Optical microscope (OM) image of the region of insertion of the stylet in the inflatable “sheath” of the µSEEG electrode. OM images of f front, g back side of the µSEEG electrode. h, i Magnified OM images at the tip f front and g back layers. The red arrow indicates the micro-hole arrays that interlock the 1st and 2nd PI layers. j Long 128-channel µSEEG electrode and comparison with a clinical electrode. k Diagram of the relative scale of human cortical neurons relative to a clinical SEEG lead and µSEEG electrodes^–. l Flexibility in manufacturing procedure to produce short 64-channel µSEEG electrodes (left) or short 32-channel µSEEG electrodes (right) and photographs showing m overall and n tip of the 64-channel µSEEG electrodes. o A perspective view of the long µSEEG electrode with partially inserted stylet illustrating the flexibility and slenderness of the electrode body.

Fig. 2. µSEEG electrodes can be used for acute and chronic implantations and recordings.

a, b Location and 3D reconstruction of possible locations of the acute and chronic implantation of µSEEG electrode devices for recording from the rat barrel cortex. c Images of the implanted µECoG electrode (left) and the µSEEG electrode (right). Note some contacts are outside brain tissue on the µSEEG electrode. d Example voltage responses across the µSEEG electrode (left) and the µECoG electrode (right) to whisker stimulation at different whisker locations, with the insets zoomed-in views of the voltages and high gamma power (HGP, 65–200 Hz). Green dots indicate a significantly different from 0.5 s before (which is baseline) air puff stimulation to the whisker (Wilcoxon rank-sum test) per channel and across trials. Number of trials >10. e Increasing responses as air puff stimulation is closer to the C3 and D3 whiskers as indicated by the current source density (CSD). f 32-channel µSEEG electrode for chronic rat recordings and a custom printed circuit board (PCB) with zero-insertion-force (ZIF) connector which electrically connects the device to the recording system via flexible flat cable (FFC). g 3D-printed headstage for the 32-channel µSEEG electrode. h–i Electrode location localization as visualized using histology. j Example voltage and high gamma power (65–200 Hz) responses without stimuli (baseline) versus with stimuli (whisker air puff). k Responses to air puffs at three different time points post implant in one rat. l Impedance measures at 1 kHz across multiple days and multiple rats; vertical bars are standard deviation from average values. For (d), (j), and (k), a.u. arbitrary units in z-scored voltage for the LFP and normalized High Gamma Power (HGP).

Fig. 3. Stimulated neural activity and ongoing clinically relevant neural dynamics can be recorded acutely using the short 64-channel µSEEG electrode.

a Direct electrical stimulation of the spinal cord during acute short µSEEG recording from the pig cortex. Gray bar indicates significantly different between current steps, Wilcoxon rank-sum test. b Voltage responses along the electrode depth with more responses significantly different to baseline (0.5 s before stimulation) occurring more with higher current levels (green dots, Wilcoxon rank-sum test). c Top: two-dimensional heat map of the largest voltage deflection from baseline per channel and per current step (left) and the maximum peak in beta (15–30 Hz) power oscillations (right). Bottom: time to peak voltage (left) and peak beta band power (right) after stimulation per channel and current step. Gray dotted line indicating 0.5 s after stimulation. d Acute implant of the short µSEEG electrode into V4 in an anesthetized NHP and ongoing evidence of burst suppression. e–g Acute implantation of short μSEEG electrodes into left anterior temporal lobe middle temporal gyrus (highlighted in blue) to be resected in the course of clinical treatment in two participants, HS1 and HS2 with a photograph of the implant, a three-dimensional reconstruction of each participants’ brain and the relative location of the μSEEG electrode (yellow dot) with a zoomed in inset view of the 64 channels as implanted. e Spontaneous ongoing activity with burst suppression along the electrode depth. f Auditory responses to low and high tones presented at random in sequence with varying jitter times while recording activity in the lateral temporal lobe. Green dots indicating p < 0.02 significant difference between low and high tones, Wilcoxon rank-sum test. g Differences in the responses varied across the depth of the electrode. Z-scored voltage responses at multiple channels at different depths averaged across trials. Green dots indicating p < 0.02 significant difference between low and high tones, Wilcoxon rank-sum test. For (a) and (b), a.u. arbitrary units in z-scored voltage for the LFP and normalized High Gamma Power (HGP).

Fig. 4. Single-unit activity could be recorded using long µSEEG electrodes.

a Three-dimensional reconstruction of the locations of the long µSEEG inserted at multiple depths into the parietal lobe, temporal lobe, and deeper into the tissue, with a zoomed-in view of the microelectrodes in the long µSEEG^,. b MRI with the overlaid CT (chambers above) and the three putative long µSEEG depths in the brain. c Example recording from the second depth to show single-unit spiking activity (filtered to between 300 and 6000 Hz). d, e Example single units recorded at electrode depth 1 (d) and depth 2 (e) showing overlaid waveforms and the autocorrelation of the spike times. f Single-unit activity was not observed at depth 3 which seemed to be in white matter, but possible MUA was recorded at depth 3. g Numbers of detected single units and MUA clusters. h Location-detected clusters relative to the mean spike rates for the different depths. Each dot is a cluster (which can represent single units or MUA). i Spike waveform measures of single units and MUA waveforms showing separation of events detected at depths 1 and 2 versus 3.