Abstract

Intraoperative and spatiotemporal monitoring of neuroinflammatory indices during brain-computer interface (BCI) implantation is essential for ensuring safety and efficacy of the procedure. Current biomolecular detection approaches are unable to obtain spatiotemporally resolved inflammatory profiling, which is important for guiding the placement of microelectrodes intraoperatively. This study presents an intraoperative spatiotemporal acute inflammation detector (ISAID) that harnesses droplet-based sampling and multiplexed titanium oxynitride (TiNO) plasmonic biosensing to assess local inflammation during the insertion of intracortical microelectrodes. Through freestanding sampling droplets and fine-tuned TiNO-based biosensors, the ISAID achieved precise, sensitive, and integrated sampling to biosensing for cytokine detection with a spatial resolution down to 610 μm and a fast equivalent bioassay time of 1.25 min. The proposed system also allows multiple ISAID biosensing modes, enabling both spatial inflammation mapping and multi-cytokine analysis. Quantitative analyses of inflammatory cytokines with in vivo mouse models demonstrate the accuracy and practical advantages of the system.

Introduction and Methods

Based on the cranial window of invasive BCI, established in vitro biomolecular sensing, such as chemiluminescent bioassay and enzyme-linked immunosorbent assay (ELISA), can potentially facilitate the characterization of BCI-induced brain inflammation. However, the time-consuming sample pretreatment and the offline bioassay may result in degradation and loss of targeted inflammatory biomarkers, impacting the accuracy and reliability of quantitative assessment. A prior investigation into titanium oxynitride (TiNO)-based plasmonic biosensors has demonstrated the potential of intraoperatively characterizing acute inflammatory responses and outlined the overall feature of brain cytokine expression at the tissue level using brain-surface rinsing specimens. The rhythm of acute inflammation responses caused by craniotomy and the placement of intracortical microelectrodes was observed.

Previous studies have shown that the plasmonic TiNO matrix has excellent optical tunability in a wide spectral range. The dielectric permittivity of TiNO varied between that of titanium nitride (TiN) and titanium dioxide (TiO2) by altering the degree of oxidation, benefiting a broad tunability of plasmonic properties considering the resonant conditions that excite the surface plasmon polariton. Therefore, the fine-tuned TiNO-sensing matrices combined with multiplexed biosensors provide a feasible approach to further improve the detection efficiency for high temporally resolved intraoperative inflammatory characterization. Nevertheless, it is worth noting that these temporally resolved intraoperative detection techniques are constrained by spatial resolution and they are not able to capture regional distribution of inflammatory responses.

Existing techniques for spatially resolved inflammation detection typically involve local biomolecular specimen collection and analysis. Methods based on in situ cerebrospinal fluid (CSF) sampling using microdialysis can provide, to some extent, local characterization of implantation-induced inflammatory responses. However, these embedded CSF samplers can impose potential risk of inducing further local trauma and thus additional inflammatory responses, which may obscure the primary inflammatory responses, leading to misinterpretation and large errors. Non-invasive biomolecular samplers with droplet-based microfluidics, however, can eliminate these limitations with the added benefit of achieving spatially resolved biomolecular specimen collection. However, the pressure exerted on the cortical surface for delivering liquid samples is a major limitation. Therefore, it is necessary to develop a droplet-based sampler with freestanding and soft-contacting features for local inflammatory characterization and profiling.

To this end, an intraoperative spatiotemporal acute inflammation detector (ISAID) is proposed herein to evaluate the local inflammation responses induced by the BCI-neuroelectrode implantation. A freestanding droplet-based sampling system is developed for soft-contact CSF cytokine collection, which minimizes the impact on brain tissue and enhances measurement accuracy and spatial resolution. A TiNO-based multiplexed plasmonic biosensor was utilized for rapid and sensitive detection of the collected droplet samples for quantitative multi-target analyses. The proposed ISAID with enhanced spatial and temporal resolution for intraoperative biomolecular sensing offers an appealing approach for rapid, closed-loop feedback during the implantation of intracortical microelectrodes.

 

Key Results and Conclusions

Thus far, inflammation assessments are mainly based on offline assessment using blood or CSF biomarkers. It often fails to offer immediate feedback, preventing the timely adjustment of the BCI implantation process. Although certain invasive techniques (e.g., microdialysis-based optical or electrochemical bioassay) have been developed to characterize the inflammatory responses in vivo, these methods do not objectively reflect the inflammation conditions induced by the neuroimplants. In addition, these methods only permit detection in single-site rather than multi-site monitoring, which prevents the achievement of spatial inflammatory mapping. In contrast, the proposed intraoperative ISAID bioassay employing soft-contact droplet sampling has diminished external buffer impact and minimized secondary damage to the local brain, thereby guaranteeing an accurate assessment of native inflammatory responses triggered by implantation. Moreover, the non-penetrating nature of ISAID avoids permanent intracerebral implantation and enables spatial mapping of cortical inflammatory responses with a spatial resolution of 610 μm. In addition, the rapid sampling-to-biosensing strategy offered by ISAID allows real-time monitoring of fast-responding yet labile inflammatory biomarkers in biofluid, positioning it as a potential online bioassay for intraoperative guidance in the placement of intracortical microelectrodes.

The implementation of the proposed ISAID system in mouse models revealed the molecular patterns of acute inflammatory responses, including the decay lifetime of inflammation and the propagation range triggered by the placement of microelectrodes. Neuroelectrode implantation also caused individual variability in neuroinflammation. The variability in inflammatory response, influenced by differences in physiological reactions and neuroelectrode implantation techniques, can result in different efficacy for long-term signal acquisition. Based on the spatiotemporal inflammatory profiling provided by ISAID, inflammatory focuses can be identified with a heightened cytokine level compared to that of the surroundings, thereby facilitating the distinction of the origin of inflammation. If the inflammation responses peaked near the implantation site, it mainly resulted from the implantation, whereas a distant inflammation center necessitates evaluation of other inflammatory triggers (e.g., surgical anomalies). The proposed ISAID, utilizing rapid intraoperative molecular detection and multiplexed inflammatory assessment, shows potential to provide closed-loop feedback during the implantation to enhance neuroelectrode implantation outcomes. To identify specific high-inflammatory regions, the spatially resolved bioassays provide an objective basis for microimplantation planning with minimized potential risks. Furthermore, the implantation strategy can be optimized by implementing ISAID to assess different implantation conditions. As demonstrated in Figure S22, an increase in the neuroelectrode insertion speed from 1 to 10 mm s−1 resulted in a reduction in the inflammatory responses. In addition, comprehensive evaluations of multiple biomarkers were also enabled with ISAID, providing a comprehensive assessment of the inflammation variation and prospective indicators for long-term physiological management. 

It should be noted that the proposed ISAID system does have certain limitations. The droplet-based sampling scheme on the brain surface yields ambiguous depth information regarding inflammatory conditions, despite the biospecimen being acquired through diffusion and exudation within the brain tissue. Additionally, the multi-channel bioassay may slightly compromise biosensing accuracy, as the temporal offset existed between signal recordings in the four micro-channels. The actual detecting signal may cease to increase during the final non-recording period, potentially leading to an overestimation of the inflammatory response in the measured signal. This shortage may be alleviated by decreasing the duration of each recording period and augmenting the cyclic numbers.

 

This study details the development of ISAID, which combined the soft-contact droplet-based sampler with multiplexed phase-sensitive biosensors to profile the acute neuroinflammatory responses during intracortical microelectrode implantation. The droplet-based sampling technique enabled minimally invasive sampling on the cortical surface, achieving high spatial resolution down to 610 μm. Simultaneously, the sensitive multiplexed biosensing unit based on the optical-tunable TiNO plasmonic matrix enhanced the efficiency for intraoperative biomolecular assessment, decreasing the equivalent detection time to 1.25 min and enabling real-time profiling of neuroinflammatory responses during neuroelectrode implantation. The ISAID system, which demonstrated two multiplexed biosensing modes for the analysis of multiple specimens or cytokines, was implemented to characterize neuroimplantation-induced acute inflammation in the mouse model. Our findings indicated that the transient inflammatory responses exhibited a distinct spatiotemporal distribution following the implantation of neuroelectrodes. The inflammatory spreading characteristics of the mouse brain revealed a decay of 13.9 min and a diffusion distance of 2.06 mm. Furthermore, ISAID is capable of concurrently assessing multiple inflammatory biomarkers at designated locations, particularly those exhibiting high inflammatory responses. This ISAID sampling-to-biosensing strategy lays the groundwork for an intraoperative inflammation early-warning system, facilitating closed-loop feedback during intracortical neuroelectrode implantation to improve surgical biosafety and prognosis.

 

 

Fig. 1. The proposed droplet-based sampling within ISAID for spatially resolved biospecimen collection. (A) Diagram of the droplet-based CSF specimen sampling on the cortical surface and continuous delivering for online inflammatory biomarker detection with the multiplexed TiNO biosensors.(B) The design of the freestanding droplet-based sampler. Scale bar, 10 mm.(C) The normal distribution of the droplet diameter during the sampling process on the abiotic surface. Scale bar, 200 μm.(D) The normal distribution of the droplet diameter during the sampling process on the mouse cortical surface. Scale bar, 500 μm.(E) The correlations between the contact angle and droplet diameter for various droplet volumes based on theoretical calculations (dashed lines) and the correspondence between the contact angle and sampling droplet diameter measured on the mouse cortical surface with a fitted regression curve (solid line).(F) The volume fluctuations recovered at different sampling rates. The volume of the total input buffer sample was 20 μL.

 

 

Fig. 2. TiNO-based plasmonic multiplexed biosensor in the ISAID system for rapid quantitative inflammatory cytokine analysis. (A) The configuration of multiplexed biosensor.(B) The TiNO biosensing chip with plasmonic resonance wavelength regulated in each sensing block. Scale bar, 10 mm.(C) Diagram of the oxidation-mediated stoichiometric regulation of the plasmonic TiNO nanofilms.(D) The TiNO plasmonic resonance wavelengths at 560, 590, 620, and 650 nm regulated by laser oxidization.(E) The XPS analyses of the Ti 2p spectra toward the two TiNO nanofilms with plasmonic sensing wavelengths at 560 and 650 nm, respectively.(F) The time-division multiplexed bioassay of different specimens in the multi-channels of the biosensor. Scale bar, 10 mm. The microfluidic channels were cyclically switched perpendicular to the optical path to acquire interferometric responses in each chamber in sequences. Iterative channel detection was recorded in real time. Multi-channel real-time phase responses were returned to the biosensing channels based on the cyclic switching patterns.(G) The phase responses of the four multiplexed micro-channels when detecting 10 ng mL−1 IL-6 simultaneously.(H) The phase responses of the multi-channel bioassay when detecting IL-6 with various concentrations at 10, 100, 1,000, and 10,000 pg mL−1. Data are represented as mean ± SD of the phase responses.(I) The regression curve for the quantitative IL-6 detection of concentrations from 1 fg mL−1 to 100 ng mL−1. Data are represented as mean ± SD, n = 3.

 

 

Fig. 3. Multi-cytokine analysis with the multiplexed plasmonic ISAID biosensor. (A–D) (A) The tandem micro-channels functionalized with four different antibody receptors for multi-cytokine analysis. Scale bar, 10 mm. The regression curves for quantitative analysis of (B) TNF-α, (C) IL-12, and (D) IL-1β, respectively, derived from triple-tested standard specimens with concentrations ranging from 1 fg mL−1 to 100 pg mL−1. Data are represented as mean ± SD, n = 3.(E) The phase responses of the four channels in the detection of the standard 10 pg mL−1 IL-6 sample. The response of the PBS buffer indicated the background response of the blank measurement.(F) The retrieved concentrations of IL-6, TNF-α, IL-12, and IL-1β based on the cross-detections of each cytokine.

 

 

Fig. 4. Spatiotemporal biomolecular analysis with integrated ISAID system. (A) The configured ISAID platform features a droplet-based sampler and a multiplexed TiNO-based biosensing system.(B) The concentrations of IL-6 detected at various locations on the surface of the hydrogel diffusion model. The scale bar is set at 3 mm.(C) The concentrations of IL-6 detected at various locations on the surface of the hydrogel model with disrupted diffusion feature, where the IL-6 molecular diffusion was confined to the left piece of hydrogel. Scale bar, 3 mm.

 

 

Fig. 5. Intraoperative inflammatory assessments with the proposed ISAID system during intracortical microelectrode implantation. (A) Implantation of microelectrodes in the MO area in the cortex of the mouse brain with robotic-controlled tungsten microneedles. Scale bar, 1 mm.(B) The IL-6 concentrations caused by single- and multi-implantation. Data are represented as mean ± SD, n = 4. p ≤ 0.05.(C) The diagram of droplet-based sampling on the mouse cortex surface for spatially resolved neuroimplantation inflammation assessments.(D) The multiplexed phase responses of ISAID for quantitative analysis of IL-6 in four samples collected from different brain locations.(E) The retrieved differential phase responses of the four micro-channels based on the multiplexed bioassay.(F) The IL-6 concentrations detected at various distances from the implantation site. Three mice under the same implantation condition were considered.(G) Comparison of the IL-6 concentrations detected within and beyond 3 mm from the implantation site. Error bars: SD. Group “Within 3 mm” (n = 10), “Beyond 3 mm” (n = 17). p ≤ 0.001.(H) The spatiotemporal mapping of the inflammatory response based on the IL-6 detection at different sites (n = 36).(I) The spatial and temporal characteristics of the acute inflammation: the propagation distance (green line) and decay lifetime (blue line) following neuroelectrode implantation were evaluated using ISAID-based IL-6 mapping.

 

 

Fig. 6. Assessing individual differences in neuroimplantation inflammation using the ISAID. (A) The IL-6 levels detected at four predetermined distances where the distant spot intersected with the edge of the cranial window.(B) The IL-6 levels detected along the radial direction from the edge of the cranial window.(C) An atypical instance demonstrating that the IL-6 level surged at a distant spot in the VIS area.(D and E) The concentrations of the four cytokines, i.e., IL-6, TNF-α, IL-12, and IL-1β, were quantified at (D) the implantation spot and (E) the abnormally high inflammation spot.

 

https://www.sciencedirect.com/science/article/pii/S2590238525003054