Quantifying physical degradation alongside recording and stimulation performance of 980 intracortical microelectrodes chronically implanted in three humans for 956-2246 days

Abstract

Motivation: The clinical success of brain-machine interfaces depends on overcoming both biological and material challenges to ensure a long-term stable connection for neural recording and stimulation. Therefore, there is a need to quantify any damage that microelectrodes sustain when they are chronically implanted in the human cortex.

Methods: Using scanning electron microscopy (SEM), we imaged 980 microelectrodes from Neuroport arrays chronically implanted in the cortex of three people with tetraplegia for 956-2246 days. We analyzed eleven multi-electrode arrays in total: eight arrays with platinum (Pt) electrode tips and three with sputtered iridium oxide tips (SIROF); one Pt array was left in sterile packaging, serving as a control. The arrays were implanted/explanted across three different clinical sites surgeries (Caltech/UCLA, Caltech/USC and APL/Johns Hopkins) in the anterior intraparietal area, Brodmann's area 5, motor cortex, and somatosensory cortex.Human experts rated the electron micrographs of electrodes with respect to five damage metrics: the loss of metal at the electrode tip, the amount of separation between the silicon shank and tip metal, tissue adherence or bio-material to the electrode, damage to the shank insulation and silicone shaft. These metrics were compared to functional outcomes (recording quality, noise, impedance and stimulation ability).

Results: Despite higher levels of physical degradation, SIROF electrodes were twice as likely to record neural activity than Pt electrodes (measured by SNR), at the time of explant. Additionally, 1 kHz impedance (measured in vivo prior to explant) significantly correlated with all physical damage metrics, recording, and stimulation performance for SIROF electrodes (but not Pt), suggesting a reliable measurement of in vivo degradation.We observed a new degradation type, primarily occurring on stimulated electrodes ("pockmarked" vs "cracked") electrodes; however, tip metalization damage was not significantly higher due to stimulation or amount of charge. Physical damage was centralized to specific regions of an array often with differences between outer and inner electrodes. This is consistent with degradation due to contact with the biologic milieu, influenced by variations in initial manufactured state. From our data, we hypothesize that erosion of the silicon shank often precedes damage to the tip metal, accelerating damage to the electrode / tissue interface.

Conclusions: These findings link quantitative measurements, such as impedance, to the physical condition of the microelectrodes and their capacity to record and stimulate. These data could lead to improved manufacturing or novel electrode designs to improve long-term performance of BMIs making them are vitally important as multi-year clinical trials of BMIs are becoming more common.

Figure 1 |. SEM images of devices and implant locations

Scanning electron microscopy was used to captured the physical damage of each array. (A) Scanning electron microscopy images were obtained at the Kavli Nanoscience Institute at Caltech and University of Utah NanoFab. (B) Each electrode was fixed to the “chuck” with carbon tape and wire bundles grounded. (C) Whole array image of platinum tipped array (M1b) from the JHU/APL participant, explanted after 2.6 years. (D) Whole array image of sputtered iridium oxide film (SIROF) tipped array (S1a) from the JHU/APL participant, explanted after 956 days. (E) NeuroPort image from the manufacturer, Blackrock Neurotech, Inc. Not shown, but two reference wires were attached to the pedestal for implantation near the array. (F) Channel layout of connected and disconnected electrodes for each type of array, and location of the wire bundle. (G) Implant locations for ten implanted arrays. Full surgical implant details and procedures can be found at their respective publications.

Figure 2 |. Electron microscopy to categorize electrode condition (end of study)

Scanning electron microscope imagery was captured of each electrode at three focal zoom strengths (500x, 1250x, 2500x) with a standard image (SE) and backscatter (BSE). Backscatter images revealed metallization of the electrode tips while standard images captured all material. (A) Humans scored each electrode by visual inspection of the images on a scale from 1 (intact) to 4 (damaged) on five metrics: metallization, insulation, shaft, growth, and separation. (B) Example heatmaps scored by one reviewer for each damage metric; blue refers generally to an intact electrode, while red indicates damage. Disconnected electrodes are marked with diagonal lines. Shanks which were completely broken off were excluded from the analysis (marked in grey). The heatmaps correspond to a platinum array at 1988 days (array AIP(a)), a platinum array at 956 days (M1(b)), and a SIROF array at 956 days (S1(a)).

Figure 3 |. Functional assessment of electrode quality (at end of study)

Prior to explant, several measurements were collected to assess functionality of each electrode’s condition: impedance, RMS noise, and signal-to-noise ratio. These data are presented from the electrode array AIP(a), M1(b), and S1(a), as identified in Figure 1. Disconnected electrodes are marked with diagonal lines and grey boxes are for electrode shanks with no data. (A) Example heatmaps of a Pt array at +5 years of implantation. Impedance is shown from 200 – 800 kΩ for the platinum arrays (AIP(a) and M1(b)) and from 30 – 250 kΩ for the SIROF array (S1(a)). (B) Example heatmaps of RMS noise for the example arrays. We plotted RMS noise from 0 – 30 μV. Greater than 30μV was shown in red. (C) Signal-to-noise ratio as a heatmap for the example arrays. SNR ranged from 0 – 5+. (D) We characterized each electrode as “able to generate a percept,” “never generated a percept,” or “generated a percept in the past, but unable to at end-of-life.” Platinum and disconnected electrodes were never stimulated, thus have no data.

Figure 4 |. Heatmap of all electrodes and metrics (at end-of-study)

Complete spatial maps for all arrays analyzed. Channel numbering starts from the top left corner, proceeding left to right, top to bottom. The wire bundle was connected to each array from the top of this layout. Scale bars indicating range of [kOhms] for impedance, μV for RMS noise, as well as average damage metrics, visual assessed from SEM images by experts. For all scale bars, blue refers generally to an intact electrode, while red indicates damage. Grey boxes have no data for Impedance (|Z|), RMS Noise, SNR, and Stim. Broken shanks are indicated with grey boxes for the visually-assessed damage metrics.

Figure 5 |. Degradation characteristics

Electrode scores for each metric visually scored by human experts were grouped according to implant duration (0 days: control, ~1000 days, ~2000 days), stimulation vs. disconnected, or tip material (Platinum vs. SIROF). Plots show the distribution of scores (violin plot), median score (white circle) and quartiles (black line). Significance was assessed via non-parametric two-sample bootstrapped confidence interval comparison and is indicated by black bars. Additional displays of array-specific information is provided in Figure S1. (A) Degradation on the platinum electrodes over time was observed for most metrics. Shaft damage was more prevalent on electrodes implanted for only ~1000 days (JHU/APL participant) than those implanted for ~2000 days (Caltech participants). (B) We observed significantly higher rates of damage to the metallization, insulation, and shaft on the SIROF shanks than then Pt shanks. Shaft damage was highest on the shanks implanted in the JHU/APL participant (all electrodes 956 days). (C) For SIROF electrodes implanted in the same participant for the same duration (2.6 yrs), we compared observed damage on electrodes through which stimulation was delivered and disconnected electrodes (see Figure 1G for spatial organization of SIROF electrodes). Two of the five damage metrics (insulation and growth) were statistically significantly worse for electrodes that had delivered stimulation vs. those that were discontinued. (D) Detailed histogram comparing all groups of electrodes. Mean and standard deviation are illustrated as a line and circle above the histogram. Further breakdown by array can be found in Figures S1 and S2.

Figure 6 |. Tip metal damage

We observed and quantified two types of tip metal degradation: Cracking/Flaking (depicted in blue in the graphs) and Pockmarked (in red). (A) Pockmarked electrode tips were present on most arrays, but most prevalent on SIROF tips. Most arrays had more than 80% of electrodes with some metal damage. (B) When comparing electrodes which were stimulated to those which were disconnected, we can observe stimulation-specific effects. Pockmark degradation features were significantly more present on stimulation electrodes vs. disconnected ones. Stimulation electrodes were also more likely to have metal damage than the disconnected ones. (C) The total charge delivered through each damaged electrode in the SIROF arrays (S1(a), S1(b), S1(c), combined) is plotted in milliCoulombs (mC). The type of tip metal degradation was not correlated with the total amount of delivered charge. (D) SEM backscatter image of a SIROF electrode with cracked/flaking metal. See Figure S3 for more examples. (E) SEM backscatter image of a SIROF electrode showing pockmarked metal degradation.

Figure 7 |. Analysis of observed damage

We evaluated relationships between damage metrics (metallization, insulation, shaft, growth, separation) and end-of-study metrics (SNR, impedance, noise). Filled in circles indicate statistical significance (p < 0.05). (A) Using Moran’s I, a measure of auto-spatial correlation, most metrics we observed were spatially correlated, suggesting damage was not due to implant/explant procedure but rather biological breakdown or tissue interaction. (B) We evaluated Spearman’s rho correlation between damage metrics and end-of-study metrics. Impedance significantly correlated with all damage metrics for SIROF electrodes, but was weakly correlated on Pt electrodes. The total charge delivered via stimulation did not correlate significantly with any of the damage metrics. (C) All damage metrics were positively correlated together, suggesting that no one metric occurred in isolation. Metal and separation metrics had the highest correlation for all groups (r > 0.75), while growth was positively correlated for SIROF electrodes for all other groups. This suggests a possible mechanistic driver for degradation for these types of electrodes. (D) SIROF electrodes had a significantly positive correlation between: impedance and RMS noise, Impedance and SNR, and RMS noise and SNR. Pt electrodes explanted at 956 days had a significantly negative correlation between Impedance and SNR. Pt electrodes explanted at 2246 days had a significantly negative correlation between Impedance and SNR, and a significantly positive correlation between RMS Noise and SNR. The total charge delivered via stimulation did not correlate with any of the other end-of-study measurements. Platinum and SIROF were oppositely correlated between impedance and SNR, the optimal impedance that correlates with detecting single units lays somewhere in the upper range of the SIROF electrodes and in the lower range for platinum, in a process that could involve tip materials and desinsulation length

Figure 8 |. Electrode functionality

Longitudinal measures of electrode health were collected throughout the duration of the implantation and directly prior to explant. (A) As expected, measured impedance was significantly higher for both groups of Pt electrodes than SIROF at end-of-study. RMS noise was significantly higher for SIROF as well as recording performance, measured by the signal-to-noise. SNR was measured as the ratio of the mean peak-to-peak waveform to its variance. (B) For each of the SIROF arrays, the majority of the electrode were able to evoke a sensation at some point during its lifespan (green + red). A small number of electrodes stopped evoking sensations prior to explant (red). (C) Impedances for all electrodes were initially higher than their specified ranges immediately after implant. They quickly settled into their respective operating ranges within the 100 days. (D) Noise was assessed via an RMS calculation and remained relatively stable after the first 200 days post-implant. Values generally decreased with time, with occasional large outliers. These outliers often corresponded to noise in the signal, degradation of filament in the analog cables or other sources of environmental noise. Shaded error bars refer to 1 standard error of the mean of all electrodes on given array.