Neural interface device

Abstract

In variants, the system can include: a controller, a set of scaffolds, and a tissue interface. The system can optionally include and / or be used with an external device. In a specific example, the system can be or include an intracortical brain-computer interface that can be embedded in-situ on the surface of the brain.

Description

NEURAL INTERFACE DEVICECROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of US Provisional Application number 63 / 606,381 filed 05-DEC-2023, which is incorporated in its entirety by this reference.

[0002] This application is related to US Application number 18 / 740,854 filed 12-JUN-2024, which is a continuation of US Application number 18 / 219,366 filed 07- JUL-2023, which claims the benefit of US Provisional Application number 63 / 359,100 filed 07-JUL-2022, US Provisional Application number 63 / 433,130 filed 16-DEC- 2022, and US Provisional Application number 63 / 438,947 filed 13-JAN-2023, each of which is incorporated in its entirety by this reference.TECHNICAL FIELD

[0003] This invention relates generally to the brain-computer interface field, and more specifically to a new and useful system and method in the brain-computer interface field.BRIEF DESCRIPTION OF THE FIGURES

[0004] FIGURE 1 is a schematic representation of a variant of the system.

[0005] FIGURE 2 is a schematic representation of an example of the system.

[0006] FIGURE 3A depicts a specific example of a proximal subsystem, including a set of scaffolds, a set of spacers, and a tissue interface.

[0007] FIGURE 3B depicts a specific example of a proximal subsystem, including a scaffold connector connecting scaffolds via conductive elements (e.g., copper bumps).

[0008] FIGURE 4 depicts an example of a proximal subsystem anchored to a tissue of interest.

[0009] FIGURE 5A depicts an example of a proximal subsystem anchored to a tissue of interest and a distal subsystem anchored to a skull, where the distal subsystem is exposed.[ooio] FIGURE 5B depicts another example of a proximal subsystem anchored to a tissue of interest and a distal subsystem anchored to a skull, where the distal subsystem is covered by skin.

[0011] FIGURE 6A depicts an example of a set of scaffolds with spacers between pairs of scaffolds.

[0012] FIGURE 6B depicts an example of a set of scaffolds with a scaffold connector connecting scaffolds via conductive elements (e.g., copper bumps).

[0013] FIGURE 7A depicts a specific example of a set of scaffolds, showing the left side of each scaffold, showing the left side of each scaffold containing an array of excitation elements (e.g., pLEDs).

[0014] FIGURE 7B depicts a specific example of a set of scaffolds (e.g., the set of scaffolds shown in FIGURE 7A), showing the right side of each scaffold containing an array of sensor elements (e.g., electrodes).

[0015] FIGURE 8 depicts an illustrative example of a proximal subsystem.

[0016] FIGURE 9 depicts an illustrative example of a distal subsystem.

[0017] FIGURE 10 depicts an example of a proximal subsystem, including a controller (e.g., a proximal controller module) and a set of scaffolds.

[0018] FIGURE 11 depicts an example of a set of scaffolds.

[0019] FIGURE 12 depicts an example including a set of scaffolds and a tissue interface.

[0020] FIGURE 13 depicts another example including a set of scaffolds and a tissue interface.

[0021] FIGURE 14 depicts an example of an implanted system.

[0022] FIGURE 15 depicts an image of an example of a tissue interface, including cells seeded into a set of scaffolds.DETAILED DESCRIPTION

[0023] The following description of the embodiments of the invention is not intended to limit the invention to these embodiments, but rather to enable any person skilled in the art to make and use this invention.1. Overview.

[0024] As shown in FIGURE 1, the system 10 can include: a controller 100, a set of scaffolds 200, and a tissue interface 300. The system 10 can optionally include and / or be used with an external device 400. However, the system 10 can additionally or alternatively include any other suitable components.

[0025] In variants, the system 10 (e.g., an intracortical brain-computer interface) can function to transmit information to a user’s brain and / or receive information from a user’s brain. In specific examples, the system 10 can perform motor control, speech decoding, sensory restoration, sensory augmentation, a combination thereof, and / or other functions.2. Examples.

[0026] In an example, the system can include an intracortical brain-computer interface that can be embedded in-situ on the surface of the brain. In a specific example, the system can include an implant, wherein the implant includes: a controller, a set of stacked scaffolds, and a set of engineered cells seeded (e.g., in 3 dimensions) between adjacent pairs of scaffolds. For example, the set of stacked scaffolds can be dies arranged in parallel layers with spacers between adjacent pairs of dies. Each scaffold can include an integrated circuit and an array of signal elements, including excitation elements (e.g., pLEDs) that can excite the engineered cells and / or sensor elements (e.g., receiving electrodes) that can record signals (e.g., action potentials) from the engineered cells. In a specific example, each scaffold can include excitation elements on the left face of the scaffold (e.g., interfacing with a first group of engineered cells) and sensor elements on the right face of the scaffold (e.g., interfacing with a second group of engineered cells). In an example, the set of engineered cells can include optogenetically modified neurons (e.g., genetically modified to express transgenic proteins sensitive to the wavelength of light emitted by the pLEDs). The bodies of the engineered cells can optionally be retained in a hydrogel, wherein the engineered cells can project axons and / or dendrites through the hydrogel to interface with the native brain tissue. In an illustrative example, the excitation elements can emit light, wherein recipient engineered cells can stimulate the native brain tissue in response to detecting the light. In another illustrative example, the engineered cells can be stimulated by native brain tissue signals, wherein sensorelements can detect associated signals in the engineered cells in response to the stimulation.3. Technical advantages.

[0027] Variants of the technology can confer one or more advantages over conventional technologies.

[0028] First, conventional intracortical brain-computer interfaces are hindered by a tradeoff between increasing bandwidth (e.g., channel count) and reducing damage to the brain; for conventional intracortical brain-computer interfaces, as bandwidth increases, the number of damaged brain cells also increases due to the increased implant size. Variants of the technology can provide a scalable intracortical brain-computer interface with a large number of channels (e.g., at least 1000, at least 5000, at least 10000, at least 50000, etc.) that can mitigate damage to the brain. In an example, the system includes a set of stacked scaffolds seeded in three dimensions with a high-density of engineered cells between the scaffolds. The engineered cells (e.g., engineered neurons) can grow projections to interface with the native cortex tissue, enabling the system to receive and / or send signals to the brain with minimal or no damage to the native neurons.

[0029] Second, variants of the technology can restore and / or augment one or more senses of a user, treat neurological diseases and / or disorders (e.g., amyotrophic lateral sclerosis, epilepsy, Alzheimer’s disease, stroke, Parkinson’s disease, motor impairments, etc.), perform motor control, perform speech decoding, and / or otherwise interface with a user’s brain. In an example, the system can both record and stimulate engineered cells interfacing with neurons in the brain (e.g., recording from the same cells that are being simulated). In an illustrative example, the system can receive information associated with brain dynamics (e.g., detecting a seizure), and emit signals to the brain based on the received information (e.g., triggering stimulation to treat the seizure). In another illustrative example, an artificial intelligence model (e.g., reinforcement learning model) can be integrated into the brain via the intracortical brain-computer interface (e.g., the model can ingest signals received from the brain via the implant, and can send model outputs to the brain via the implant).

[0030] Third, conventional methods of interfacing with neurons require large capacitive electrode systems with cumbersome hermetic feedthroughs (to sealimplanted electrodes). In variants, this technology can provide content to the brain (e.g., via RGCs, via neurons, etc.) using a display implant that has a higher resolution (e.g., a 1:1 mapping of cells to light emission systems on the display), a higher number of signal elements, and / or a smaller form factor with fewer or no hermetic feedthroughs. In an example, variants of the technology can be scalable (e.g., to thousands of pixels, to hundreds of thousands of pixels, etc.).

[0031] However, further advantages can be provided by the system and method disclosed herein.4. System.

[0032] As shown in FIGURE 1, the system 10 can include: a controller 100, a set of scaffolds 200, and a tissue interface 300. The system 10 can optionally include and / or be used with one or more of: an external device 400, a computing system (e.g., local, remote, distributed, etc.), a database (e.g., a system database, a third-party database, etc.), user interface, and / or any other components. The computing system can include one or more: CPUs, GPUs, TPUs, custom FPGA / ASICS, microprocessors, servers, cloud computing, and / or any other suitable components. The computing system can be local, remote (e.g., cloud computing server, etc.), distributed, and / or otherwise arranged relative to any other system or module. The user interface can receive one or more inputs (e.g., from a user), display one or more outputs (e.g., processed and / or unprocessed data received from the sensor elements, model outputs, etc.), display any other parameters, and / or otherwise function.

[0033] The system 10 can be used with a tissue of interest of a user. For example, the system 10 can interface with the tissue of interest. In a specific example, the system 10 can be physically coupled to the tissue of interest (e.g., via tissue anchors). In another specific example, the system 10 can send signals to and / or receive signals from the tissue of interest (e.g., via the tissue interface). The tissue of interest preferably includes a brain (e.g., brain surface, cerebrum, cortex, etc.), but can alternatively include a brain stem, eye, spinal cord, ear, muscle, skin, nerve, and / or any other body region of the user. The user can be a human, an animal (e.g., rabbit, chimpanzee, mouse, rat, etc.), and / or any other organism. The system 10 and / or components therein are preferably implanted at one or more implantation locations (e.g., locations in, on, and / or near the tissue of interest). Examples of implantation locations includelocations within or on the surface of: the brain, dura mater, pia mater, cortex, skull, skin, muscle, and / or any other implantation location.

[0034] As used herein, the z-axis is defined as extending away from the tissue of interest (e.g., away from the brain surface and towards the skull). As used herein, the x-axis is defined as orthogonal to a face of a scaffold (e.g., towards an adjacent scaffold). In an example, the height of a scaffold is along the z-axis, the thickness of a scaffold is along the x-axis, and the width of a scaffold is along the y-axis. The coordinates, as used herein, are intended only as a reference and are not intended to restrict the orientation of the system 10 relative to a global coordinate system. The system 10 can be arranged in any orientation (e.g., where the x-axis corresponds to a vertical axis, where the y-axis corresponds to a vertical axis, where the z-axis corresponds to a vertical orientation, etc.).

[0035] The system 10 can optionally include multiple subsystems. An example is shown in FIGURE 2. For example, the system 10 can include a proximal subsystem 12 implanted proximal to the tissue of interest (e.g., in contact with the tissue of interest) and a distal subsystem 14 implanted distal to the tissue of interest (e.g., at or near the skin surface, crossing the skin surface, crossing the skull, etc.). In an example, the proximal subsystem 12 (e.g., intracranial subsystem) can include a proximal controller module 120 (e.g., intracranial controller module), the set of scaffolds 200, the tissue interface 300, and / or any other system components. Examples are shown in FIGURE 3A, FIGURE 3B, and FIGURE 10. In an example, the distal subsystem 14 (e.g., an extracranial subsystem) can include a distal controller module 140 (e.g., extracranial controller module) of the controller 100 and / or any other system components. In an example, the extracranial subsystem can be implanted partially or completely external to the skull (e.g., at or near the skin surface, passing through the skin surface, etc.). An example is shown in FIGURE 9. In a first specific example, the extracranial subsystem can cross the skin surface (e.g., the extracranial subsystem is exposed outside the skin). In a second specific example, the extracranial subsystem can be implanted beneath the skin surface (e.g., where the skin is sutured to cover the extracranial subsystem). In an example, the intracranial subsystem can be implanted partially or completely internal to the skull. An example is shown in FIGURE 8. In a first specific example, a section of the skull is removed, and the intracranial subsystemextends from the surface of the brain to a location (e.g., height) within (e.g., at or below) the pre-removal location of the inner surface of the section of the skull. In an illustrative example, the intracranial subsystem is located within between the brain and the inner surface of the skull (e.g., beneath the skull). In a second specific example, a section of the skull is removed, and the intracranial subsystem extends from the surface of the brain to a location (e.g., height) outside (e.g., above) the pre-removal location of the inner surface of the section of the skull. In an illustrative example, the intracranial subsystem passes partially or completely through the skull. However, subsystems can be otherwise configured.

[0036] The system 10 can optionally include a housing around all or a portion of the system 10 and / or components therein (e.g., the proximal subsystem 12 and / or the distal subsystem 14). The housing can be: a coating (e.g., an encapsulant), a wrapping, a jacket, a cap, and / or any other material covering all or a portion of the system 10 and / or components therein. Examples of housing materials can include: silicone, polyether-ether-ketone (PEEK), metals (e.g., titanium), a combination thereof, and / or any other suitable material. The housing(s) can function to seal system components, provide flanges for anchoring, facilitate implantation, and / or perform other functions. In a first specific example, the proximal subsystem 12 can include a housing (e.g., a silicon jacket) that includes one or more flanges, wherein the flange(s) are used as anchoring points for one or more tissue anchors (e.g., brain anchors). In a second specific example, the distal subsystem 14 can include a housing that includes one or more flanges, wherein the flange(s) are used as anchoring points for one or more skull anchors. In another illustrative example, the housing for the distal subsystem 14 can include a cap exposed beyond the skull. The cap can be polyether- ether-ketone (PEEK), metal (e.g., titanium), and / or any other suitable material.

[0037] The system 10 and / or components therein can optionally be anchored (e.g., adhered, secured, etc.) to tissue (e.g., skin, bone, dura mater, pia mater, etc.) at an implantation location. Anchoring methods can include: tissue adhesives, extrusions that act as anchors, packing, anchors (e.g., tacks, pins, screws, etc.) inserted through the component (e.g., through a flange in a housing around the system 10 and / or components therein), inducing fibrotic growth in or around the system 10, anchoring via the tissue interface 300, a combination thereof, and / or any other method.Examples are shown in FIGURE 4, FIGURE 5A, FIGURE 5B, FIGURE 13, and FIGURE 14. In an illustrative example, one or more anchors can secure the proximal subsystem 12 (e.g., intracranial subsystem) to the brain tissue. In a specific example, the system 10 can include an anchor configured to secure the set of scaffolds 200 to the surface of the brain of a user. In another illustrative example, one or more bone screws can secure the distal subsystem 14 (e.g., extracranial subsystem) to the skull. In another illustrative example, sutures can secure the distal subsystem 14 to skin. The system 10 and / or components therein can optionally be sealed to tissue at the implantation location(s). Examples of sealants can include dental acrylic, silicone adhesive, and / or any other sealants.

[0038] The system 10 can optionally include a spring, which can function to bias the proximal subsystem 12 and / or other system components against the tissue of interest and / or against other tissue at an implantation location. For example, the spring can be connected to the intracranial subsystem, pressing the intracranial subsystem against the brain surface (e.g., pia mater, cortex tissue, etc.). The spring can be positioned between the proximal subsystem 12 and the distal subsystem 14, between the proximal subsystem 12 and other tissue (e.g., skull, skin, dura mater, etc.), and / or any other components. The spring can be a coil spring, a plate (e.g., cambered plate), deformable material, and / or any other spring. The spring material can include metal, plastic, a hydrogel, an elastomeric material, and / or any other material.

[0039] The controller 100 functions to control the set of signal elements 240 (e.g., excitation elements, sensor elements, etc.) and / or other electronic components of the set of scaffolds 200. For example, the controller 100 can transmit control instructions (e.g., excitation parameters, sensor parameters, etc.) to the set of scaffolds 200 and / or receive data from the set of scaffolds 200. The controller 100 can additionally or alternatively function to: transmit and / or receive information from an external device 400; process data (e.g., compress data, convert data, transform data, etc.); perform safety checks; provide power to one or more system components (e.g., to the set of scaffolds 200); regulate power; acquire data; and / or perform other functions. In a specific example, the controller can provide power to the set of scaffolds 200, wherein the controller 100 itself can be powered by an external power source and / or an onboard power source. The controller 100 and / or components therein arepreferably physically connected to the set of scaffolds 200, but can alternatively be not physically connected to the set of scaffolds 200.

[0040] In examples, the controller 100 can include one or more: power components (e.g., power source, power receiver components, power rectification and / or regulation components, etc.), communication elements, processing systems (e.g., processor, memory, data processing circuitry, logic, microcontroller, integrated circuits, etc.), filters (e.g., capacitors, resistors, etc.), controller sensors, analog-to- digital converter (ADC), ultrasonic backscatter reduction components, optical backscatter reduction components, and / or any other system component. Integrated circuits can include: field-programmable gate array (FPGA), system on a chip, application-specific integrated circuit (ASICs), and / or any other integrated circuit. Communication elements can include: receivers, transmitters, transceivers, antenna, wired connections, and / or any other components. Communication elements can be wired and / or wireless (e.g., radio frequency (RF), infrared (IR), Bluetooth, BLE, NFC, etc.).

[0041] The controller sensors can optionally function to measure a controller state, a user state, tissue state, and / or collect any other data. Examples of controller sensors can include moisture sensor (e.g., humidity sensor), temperature sensor, pose and / or motion sensor (e.g., IMU), metabolism sensor, pH sensor, current sensor, voltage sensor, ADC (e.g., to measure power), and / or any other sensor. The controller 100, external device 400, the set of signal elements 240, and / or any other components can optionally be controlled based on data received from the controller sensors. In a first example, power can be halted to one or more components in response to detection or occurrence of a predetermined controller state (e.g., indicating water ingress, temperature above a threshold, electrical shorts, etc.). In a second example, data from the set of signal elements 240 can be processed (e.g., reducing noise) based on the sensor data.

[0042] The controller 100 can optionally include one or more modules. For example, the controller 100 can include a proximal controller module 120 proximal to the tissue of interest and / or a distal controller module 120 distal to the tissue of interest. The proximal controller module 120 can function to interface with the set of scaffolds 200 and / or perform other controller 100 functions. The distal controllermodule 140 can function to process data, provide power, interface with an external device 400 (e.g., wirelessly, via a wired connection such as a USB-C, etc.), and / or perform other controller 100 functions. In an example, the distal controller module 140 can include: power components (e.g., power source, power receiver, regulators, etc.), filters, processing systems (e.g., processor, memory, data processing circuitry, logic, microcontroller, integrated circuits, etc.), controller sensors, communication elements, and / or any other controller 100 components. In an example, the proximal controller module 120 can include: power components (e.g., power receiver, regulators to filter noise and / or for other functions, etc.), filters, processing systems, controller sensors, communication elements, and / or any other controller 100 components.

[0043] In a first variant, the controller 100 can include a proximal controller module 120 (e.g., intracranial controller module) and a distal controller module 140 (e.g., extracranial controller module). For example, the distal controller module 140 can be connected to the proximal controller module 120 (e.g., via a wired connection) and optionally the external device 400; the proximal controller module 120 can be connected to the set of scaffolds 200. In specific examples, the proximal controller module 120 can be connected to the set of scaffolds 200 via: a direct bond to the scaffolds, a direct bond to a scaffold connector, a wired connection to the scaffolds, a wired connection to a scaffold connector, and / or otherwise connected. In a specific example, the distal controller module 140 can be connected to the external device 400 via a wired and / or a wireless connection. The proximal controller module 120 and the distal controller module 140 are preferably physically connected (e.g., by a cable and / or other wired connection), but can alternatively be wirelessly connected. In a specific example, the connection between the proximal controller module 120 and the distal controller module 140 can be a flexible connector. In a specific example, a wired connection between the proximal controller module 120 and the distal controller module 140 can include one or more bends in the wired connection (e.g., for strain relief, to facilitate implantation, etc.). In a specific example, the controller 100 can include a proximal controller module 120 and a distal controller module 140, wherein the proximal controller module 120 is coupled to the set of scaffolds 200, wherein the distal controller module 140 is anchored to the skull of the user, and wherein the distal controller module 140 is communicatively coupled to an external device 400.

[0044] In a second variant, the controller 100 can include (only) a distal controller module 140 (e.g., extracranial controller module). For example, the distal controller module 140 can be connected to the set of scaffolds 200 and optionally the external device 400. In specific examples, the distal controller module 140 can be connected to the set of scaffolds 200 via: a wired connection to the scaffolds, a wired connection to a scaffold connector, and / or otherwise connected. In a specific example, the distal controller module 140 can be connected to the external device 400 via a wired and / or a wireless connection.

[0045] In a third variant, the controller 100 can include (only) a proximal controller module 120 (e.g., intracranial controller module). For example, the proximal controller module i2ocan be connected to the set of scaffolds 200 and an external device 400. In specific examples, the proximal controller module 120 can be connected to the set of scaffolds 200 via: a direct bond to the scaffolds, a direct bond to a scaffold connector, a wired connection to the scaffolds, a wired connection to a scaffold connector, and / or otherwise connected. In a specific example, the proximal controller module 120 can be connected to the external device 400 via a wireless connection.

[0046] However, the controller 100 can be otherwise configured.

[0047] The set of scaffolds 200 functions to emit excitation signals and / or to detect signals from cells within the tissue interface 300. The set of scaffolds 200 can be connected to the controller 100 (e.g., to the proximal controller module 120), to the tissue interface 300, and / or to any other system component. For example, the set of scaffolds 200 can receive control instructions (e.g., excitation parameters, sensor parameters, etc.) from the controller 100, and can emit excitation signals (via excitation elements) and / or receive cell signals (via sensor elements) according to the control instructions. For example, the set of scaffolds 200 can receive control instructions (e.g., excitation parameters, sensor parameters, etc.) from the controller 100, and can emit excitation signals (via excitation elements) and / or receive cell signals (via sensor elements) according to the control instructions.

[0048] Each scaffold can include one or more of: signal elements (e.g., one or more arrays of signal elements), drivers (e.g., pLED drivers), power components (e.g., internal regulation), processing systems (e.g., integrated circuits (ICs)), registers,multiplexers, ADCs, and / or any other system component. Examples are shown in FIGURE n and FIGURE 12. Scaffolds preferably include a silicon backing, but can additionally or alternatively include any other material. In an example, each scaffold can include one or more ICs (e.g., ASICs). In a specific example, a scaffold can include a die containing an array of signal elements (e.g., an array of excitation elements and / or an array of sensor elements) and an IC (e.g., including drivers for excitation elements, internal regulation, registers, multiplexers, an ADC, etc.). In an illustrative example, each scaffold can be configured with the set of signal elements 240 on a proximal end (e.g., fin) of the scaffold configured to be proximal to the tissue of interest, and the IC on a distal end of the scaffold configured to be distal to the tissue of interest (e.g., and proximal to the controller 100). Scaffolds in the set of scaffolds 200 can contain the same or different signal elements, the same or different signal element configurations (e.g., location of the signal elements on the scaffolds), the same or different IC components, and / or can be otherwise configured. In an illustrative example, a subset of the set of scaffolds 200 can include ADCs (e.g., where connections between the scaffolds enable a scaffold without an ADC to use an ADC on a neighboring scaffold). Scaffolds can optionally ingest and / or output data using serial protocols (e.g., using Inter-Integrated Circuit, Serial Peripheral Interface, etc.). In a specific example, signals output from each scaffold (e.g., data received from sensor elements; output via an ADC on the scaffold) can be serialized across the set of scaffolds 200 and transmitted to the controller 100 (e.g., to the proximal controller module 120).

[0049] In an example, the number of scaffolds in the set of scaffolds 200 can be between 1-10,000 or any range or value therebetween (e.g., 1-5, 5-10, 10-50, 50- 100, 100-256, at least 2, at least 4, at least 8, at least 10, at least 64, at least 100, etc.). Each scaffold is preferably planar, but can alternatively be any other geometry. In an example, the thickness of a scaffold (e.g., in the x-direction) can be between 5pm- 1000pm or any range or value therebetween (e.g., 2opm-6opm, 20pm, 50pm, less than 500pm, less than 100pm, less than 20pm, etc.). In a first embodiment, the thickness is constant along the height of the scaffold (e.g., in the z-direction). In a second embodiment, the thickness can be variable. In an example, the thickness at the fin (proximal to the tissue of interest) can be less than the thickness at other locationson the scaffold. In another example, the thickness can taper at the edge of the fin (e.g., to facilitate cell loading).

[0050] The scaffolds are preferably stacked in parallel layers (e.g., at intervals along the x-axis), but can additionally or alternatively include non-parallel scaffolds (e.g., a second stack of scaffolds orthogonal to a first stack of scaffolds) and / or can be arranged in any other geometry. In an example, the footprint of the set of scaffolds 200 (e.g., on the tissue of interest; in the xy-plane) can be between 0.5mm2- 5cm2or any range or value therebetween (e.g., iomm2-ioomm2, 50mm2, etc.). In an example, the center-to-center distance between scaffolds (e.g., in the x-direction) can be between 3opm-iooopm or any range or value therebetween (e.g., 5opm-ioopm, approximately 80pm, less than 150pm, at least 50pm, at least 100pm, at least 200pm, etc.). In an example, the distance between adjacent fins (e.g., the gap between the surface of a first fin to the surface of the adjacent fin; the thickness of a spacer as described below; etc.) can be between 2opm-iooopm or any range or value therebetween (e.g., 2opm-ioopm, loopm-iooopm, approximately 500pm, at least 100pm, at least 200pm, etc.). The end faces of the stack of scaffolds can optionally be adhered to an edge support. In an example, the thickness of the edge support can be between iopm-10, ooopm or any range or value therebetween. The material of the edge support can be glass, a polymer, metal (e.g., titanium), and / or any other suitable material.

[0051] A set of spacers 220 can optionally interface with the set of scaffolds 220. For example, a spacer can be positioned between each pair of adjacent scaffolds. The material of the set of spacers 220 can be glass, a polymer (e.g., liquid crystal polymer), metal (e.g., titanium), any encapsulation material, and / or any other suitable material. The set of spacers 220 can optionally extend partially down the set of scaffolds 200 (e.g., in the z-direction; from the distal end of a scaffold towards the fin), forming an open trench between fins. In an example, the height of a spacer (e.g., in the z-direction) can be between sopm-sooopm or any range or value therebetween (e.g., 200pm- 400pm). In an example, the height of a spacer (e.g., in the z-direction) can be between io%-9O% of the height of the scaffolds or any range or value therebetween (e.g., at least 25%, at least 50%, etc.). In an example, the thickness of a spacer (e.g., thickness of the trench) can be between 2opm-iooopm or any range or value therebetween(e.g., 2opm-ioopm, loopm-iooopm, approximately 500pm, at least 100pm, at least 200pm, etc.). In an example, the depth of the trench between fins (e.g., in the z- direction) can be between 5o m-5ooo m or any range or value therebetween (e.g., 2oo m-4oo m).

[0052] Scaffolds (e.g., fins) can optionally be coated with proteins (e.g., growth factors, brain proteins, etc.), poly-d-lysine, laminin, and / or other materials. Scaffolds (e.g., fins) can optionally undergo one or more surface modifications. For example, the surface of the scaffolds (e.g., the surface of the fins) can undergo plasma activation. In variants, this can increase the hydrophilicity of the surface of the scaffold, increasing adhesion between the scaffolds and the cell support 340.

[0053] The set of scaffolds 200 can optionally include one or more scaffold connectors. Scaffold connectors can include: interscaffold connectors (e.g., interdie connectors) connecting multiple scaffolds, scaffold-controller connectors connecting one or more scaffolds to the controller 100 (e.g., to the proximal controller module 120), and / or any other connectors. In an example, scaffold connectors can include connections through scaffolds (e.g., through-silicon via) and / or edge out connectors (e.g., connections along the distal edge of the scaffolds). The scaffold connectors can optionally include an interposer layer between the controller 100 and the set of scaffolds 200.

[0054] The set of spacers 220 can optionally include a set of raised conductive elements (e.g., copper bumps), wherein the scaffold connector connects multiple raised conductive elements. Examples are shown in FIGURE 3B and FIGURE 6B. For example, a spacer between a left scaffold and a right scaffold can include a first raised conductive element in contact with the right face of the left scaffold and second raised conductive element in contact with the left face of the right scaffold; the scaffold connector can then connect the first raised conductive element to the second raised conductive element (e.g., thus connecting the left scaffold to the right scaffold). In a specific example, the face of the distal edge of each scaffold is ground down to reveal conductive elements in the scaffolds, enabling electrical connectivity between the raised conductive elements on the set of spacers 220 and the set of scaffolds 200. In an illustrative example, the scaffold connector can be a metal trace along the top edge of the set of scaffolds 200, connecting all scaffolds in the set of scaffolds 200 via theraised conductive elements. In an illustrative example, the scaffold connector can be bonded (e.g., wire bonded) to a cable connected to the controller 100 (e.g., to the proximal controller module 120 and / or to the distal controller module 140).

[0055] In an illustrative example, manufacturing the set of scaffolds 200 can include one or more of: forming individual scaffolds (e.g., thinning a full thickness silicon wafer), stacking the scaffolds with a spacer (e.g., liquid crystal polymer) between each pair of scaffolds, compressing the scaffolds, bonding the scaffolds (e.g., polymer bonding, oxide-oxide bonding, etc.; optionally including etching holes within the scaffolds to facilitate the bonding), grinding the face of the distal edge of each scaffold is to reveal conductive elements in the scaffolds, connecting the scaffolds to the controller 100 (e.g., to the proximal controller module 120), and / or any other suitable steps.

[0056] The set of signal elements 240 can include excitation elements, sensor elements, and / or any other element. Each signal element in set of signal elements 240 can interface with (e.g., record from and / or excite) 1 cell, more than 1 cell (e.g., at least 2, at least 3, etc.), o cells, a variable number of cells, a randomly determined number of cells, all cells within a trench, all cells within a trench that are less than a predetermined distance from the signal element (e.g., 100pm, 50pm, 10pm, 5pm, etc.), and / or any other number of cells. In an illustrative example, a cell in the tissue interface 300 can be stimulated via an excitation element on a scaffold, and a sensor element (on the same scaffold or an adjacent scaffold) can record signals from the same cell.

[0057] In an example, a scaffold face can include between 10-50,000 signal elements or any range or value therebetween (e.g., 1,000-10,000; 2,000-5,000; at least 10; at least 100; at least 1,000; etc.). In a specific example, a scaffold face can include between 10-100,000 sensor elements (e.g., recording electrodes) or any range or value therebetween (e.g., 1,000-10,000; 2,000-5,000; at least 10; at least 100; at least 1,000; etc.). In another specific example, a scaffold face can include between 10- 100,000 excitation elements (e.g., pLEDs) or any range or value therebetween (e.g., 1,000-10,000; 2,000-5,000; at least 10; at least 100; at least 1,000; etc.). In an example, the total number of signal elements across the set of scaffolds 200 can be between looo-ioomillion or any range or value therebetween (e.g., at least 50k, atleast look, at least 395k, at least 1 million, etc.). In an example, the diameter of a signal element can be between 1 pm - imm or any other range or value therebetween. In an example, the pitch in an array of signal elements on a scaffold (e.g., center-to-center distance between signal elements in the set of signal elements 240) can be between 5pm - 10mm or any range or value therebetween.

[0058] Signal elements can be located on a single face of a scaffold or multiple faces of a scaffold (e.g., opposing faces of a scaffold). Examples are shown in FIGURE 6A, FIGURE 6B, FIGURE 7A, and FIGURE 7B. In an example the set of signal elements 240 can include one or more arrays of signal elements on each scaffold of the set of scaffolds 200. In a specific example, the set of signal elements 240 includes one array of signal elements on the two end scaffolds (e.g., the two scaffolds flanking the set of scaffolds 200) and two arrays of signal elements on the intervening scaffolds (e.g., all scaffolds between the two end scaffolds). In a second specific example, the set of signal elements 240 includes two arrays of signal elements on all scaffolds in the set of scaffolds (e.g., where the array of signal elements on the outer face of each end scaffold is not used). In a specific example, sensor elements (e.g., an array of recording electrodes) and excitation elements (e.g., an array of pLEDs) are located on opposing faces of a scaffold. Examples are shown in FIGURE 7A and FIGURE 7B. In an illustrative example, for a trench defined by a left scaffold and a right scaffold, excitation elements are located on the right face of the left scaffold (forming the left side of the trench) and sensor elements are located on the left face of the right scaffold (forming the right side of the trench). In another illustrative example, for a trench defined by a left scaffold and a right scaffold, sensor elements are located on the right face of the left scaffold (forming the left side of the trench) and excitation elements are located on the left face of the right scaffold (forming the right side of the trench).

[0059] In an example, the system 10 can include: a first scaffold including an array of pLEDs on a first face of the first scaffold; a second scaffold including an array of recording electrodes on a first face of the second scaffold (e.g., where the first face of the first scaffold faces the first face of the second scaffold); a spacer separating the second scaffold from the first scaffold, the spacer positioned between the first face of the first scaffold and the first face of the second scaffold (e.g., where the spacer is in contact with the first face of the first scaffold and the first face of the second scaffold);a cell support (e.g., gel) positioned between the array of pLEDs and the array of recording electrodes; and a set of cells 320 (e.g., genetically modified cells) retained within the cell support, wherein the array of recording electrodes is configured to receive signals from the set of cells 320, wherein the array of pLEDs is configured to transmit light signals to the set of cells 320. In a specific example, the first scaffold further includes a second array of recording electrodes on a second face of the first scaffold, wherein the second face of the first scaffold is opposite the first face of the first scaffold, and wherein the second scaffold further includes a second array of pLEDs on a second face of the second scaffold, wherein the second face of the second scaffold is opposite the first face of the second scaffold.

[0060] A sensor element can include a light sensor system (e.g., detecting a wavelength of light emitted by cells in the tissue interface 300), an electrical sensor system (e.g., a recording electrode sensing a current signal and / or voltage signal from cells in the tissue interface 300), and / or any other signaling system. In a specific example, the sensor elements can be or include an array of recording electrodes. Sensor elements can measure signals (e.g., action potentials) from cells in the tissue interface 300. In specific examples, the sensor elements can measure the presence and / or concentration of one or more cellular molecules (e.g., calcium ions), cell response to an excitation signal, cell response to a signal from the tissue of interest, and / or any other cell response. However, sensor elements can be otherwise configured.

[0061] An excitation element can include a light emission system, an electrical emission system (e.g., emitting a current signal and / or voltage signal), and / or any other signaling system. The light emission system preferably includes a pLED, but can additionally or alternatively include laser diodes, phosphors, and / or any other light emission system. In a first specific example, the excitation elements can be or include an array of electrodes. In a second specific example, the excitation elements can be or include an array of pLEDs. The pLEDs (e.g., an array of pLEDs, multiple subarrays of pLEDs, etc.) can optionally be configured with a common cathode integration or a common anode configuration. The light parameters of the light emission system (e.g., wavelength, frequency, photon energy, intensity, flux, amplitude, etc.) preferably correspond to an optogenetic actuator (e.g., an opsin) associated with cells in the tissueinterface 300, but can alternatively not be associated with an optogenetic actuator. In an example, the wavelength (e.g., spectral peak) can be between 4oonm-8oonm or any range or value therebetween.

[0062] The excitation elements can operate according to excitation parameters (e.g., excitation instructions) received from the controller 100. Excitation parameters can include spatial parameters (e.g., which excitation elements to operate), temporal parameters (e.g., when to initiate signal emission, length of signal emission, etc.), intensity and / or amplitude parameters (e.g., intensity and / or amplitude of light), light wavelengths, and / or any other parameters defining the signals emitted by the excitation elements. For example, the emission parameters can prescribe a timeseries of light array patterns (e.g., including wavelength, intensity, spatial information, state change instructions, and / or any other parameters for each excitation element), wherein each light array pattern encodes information (e.g., a content frame). The excitation parameters are preferably determined based on content (e.g., visual data, sensory data, other external information, artificial data, any other data, etc.), but can additionally or alternatively be determined based on cell state, signal element information (e.g., the current state of each signal element in the set of signal elements 240), controller state information, calibration information, and / or any other information. For example, the emission parameters can be determined such that the resulting excitation signals collectively encode the content.

[0063] The set of scaffolds 200 can optionally include and / or be coupled to one or more optics components (e.g., microoptical components). The optics components can function to collimate light, homogenize light, focus light, reduce light backscatter, and / or otherwise modify light emitting to or from the excitation elements. Examples of optics components can include: lenses (e.g., microlens, diffractive lens, metalens, etc.), back reflectors, a modified pLED shape, waveguides, a spacing gap, a combination thereof, and / or any other optics components.

[0064] However, the set of scaffolds 200 can be otherwise configured.

[0065] The tissue interface 300 functions to convert the excitation signals from the excitation elements to neural signals that can be interpreted by the brain and / or to produce action potentials (e.g., associated with neural signals received from the brain) that can be detected by the sensor elements.[oo66] The tissue interface 300 can include a set of cells 320, a cell support 340, and / or any other suitable components. All or a portion of the tissue interface 300 (e.g., the cell support 340 of the tissue interface 300) is preferably in contact with the tissue of interest, but can alternatively be separated from the tissue of interest (e.g., separated by pia mater), and / or otherwise positioned. In an example, the set of cells 320 can interface (e.g., directly or indirectly) with the tissue of interest. In a specific example, the set of cells 320 can interface with native neurons in the brain after growing projections (e.g., axons, dinitrides, etc.) into the brain tissue. For example, the set of cells 320 (e.g., genetically modified cells) can include axons and / or dendrites that extend out of the cell support (e.g., gel), wherein the axons and / or dendrites interface with native neurons in the brain of the user. In an illustrative example, when activated by excitation signals from the excitation elements, the set of cells 320 can evoke activity in the brain. In another illustrative example, when activated by neural signals from the native brain tissue, the set of cells 320 can produce action potentials that can be detected by sensor elements.

[0067] The set of cells 320 preferably includes neurons, but can additionally or alternatively include stem cells, neural progenitor cells, and / or any other cell type. In an example, the set of cells 320 can include cells derived from pluripotent stem cells. In a specific example, the set of cells 320 can include neurons (e.g., neurons derived from pluripotent stem cells) and / or neural progenitor cells (e.g., neural progenitor cells derived from pluripotent stem cells). The set of cells 320 can be cells derived from the user or cells derived from another organism. The set of cells 320 can include genetically modified cells, unmodified cells, and / or a combination thereof. Genetically modified cells can include optogenetically modified cells with light-sensitive biochemical signaling pathways (e.g., such that the cells produce biochemical signals in response to receiving certain wavelengths of light), cells modified to overexpress collagenous (e.g., to enable the cells to penetrating residual pia mater), hypoimmune cells (e.g., genetically modified to reduce immune response due to implantation), cells modified to include a small molecule killswitch (e.g., transfected with a killswitch gene), cells modified to include inducible transcription factors to drive cell fate, cells modified to include a calcium sensor, cells modified to contain engineered cell-surface molecules (e.g., to enable the cells to form specific synaptic connections), acombination thereof, and / or other genetically modified cells. Genetic modifications can optionally be inducible. Examples of cells that can be genetically modified include: organoids, cells selected from an organoid (e.g., a specific cell type), neurons, neural progenitor cells, stem cells (e.g., wherein the stem cells are genetically modified prior to differentiation), any animal cell (e.g., human cell), and / or any other cell. In a specific example, the genetically modified cells can include genetically modified neurons (e.g., genetically modified neurons derived from pluripotent stem cells) and / or genetically modified neural progenitor cells (e.g., genetically modified neural progenitor cells derived from pluripotent stem cells).

[0068] In variants leveraging modified cells, the set of cells 320 can be genetically modified by transfecting cells with a light-sensitive protein (e.g., using a virus with a plasmid and capsid) that acts as an optogenetic actuator (e.g., optogenetic effector) and / or optogenetic sensor. Optogenetic actuators produce a biochemical signal (e.g., an action potential) in response to receiving light at a specific wavelength; optogenetic sensors produce light at a specific wavelength based on (e.g., proportional to) a state of the cell (e.g., a concentration of a given molecule). However, optogenetic actuators and optogenetic sensors can be otherwise defined. In variants, optogenetic sensors can be coupled to optogenetic actuators in a cell (e.g., a fusion construct) such that when the optogenetic actuator receives a first wavelength of light the optogenetic sensor is activated to can emit a second wavelength of light based on the cell state.

[0069] The virus used to transfect the set of cells 320 can optionally be targeted to a specific cell type (e.g., general soma cells, neurons, neural progenitor cells, stem cells, etc.). The capsid can be an adeno-associated virus capsid (e.g., AAV2.7M8) and / or any other suitable capsid. The plasmid can be an opsin, a fluorescent biosensor protein, and / or any other suitable plasmid. Opsin examples include: CheRiff, ChroMD, ChroME, ChroME2S, ChRmine (e.g., ChRmine-mScarlet) , ChrimsonR (e.g., including red-shifted variants), ReachR, and / or any other opsin. In an illustrative example, the set of cells 320 can be transfected using: AAV2.7m8 hSyni-ChRmine- KV2.I-WPRE. Fluorescent biosensor protein examples include: GCaMP8s, GCaMP8m, jRGecoia, YCaMP, iGECI, and / or any other suitable protein. The opsin can be activated by blue light, green light, red light, and / or any other wavelength. Different plasmids (e.g., with different wavelength sensitivities) can optionally be used forcontent transfer (e.g., input), sensing, (e.g., cell monitoring), and / or sensing activation. Different plasmids can optionally be used for different types of content (e.g., different communication modalities). However, any optogenetic method can be implemented.

[0070] The set of cells 320 are preferably seeded between scaffolds in the set of scaffolds 200 (e.g., in the trenches between the fins), but can be otherwise integrated into the system 10. An example is shown in FIGURE 15. For example, the set of cells 320 can be seeded between each pair of adjacent scaffolds in the set of scaffolds 200. In an example, the system 10 can include: a set of cells 320 (e.g., genetically modified cells), a set of scaffolds 200, wherein the set of cells 320 are seeded between adjacent pairs of scaffolds in the set of scaffolds 200; and a controller 100 communicatively connected to the set of scaffolds 200. In a specific example, each scaffold in the set of scaffolds 200 includes: an array of recording electrodes configured to receive signals from the set of cells 320; and an array of pLEDs configured to transmit light signals to the set of cells 320 based on control instructions; wherein the controller 100 is configured to: receive data from the array of recording electrodes; and determine control instructions for the array of pLEDs.

[0071] There is preferably greater than a 1:1 ratio of cells to signal elements (e.g., at least 2:1, at least 3:1, etc.), but can alternatively be a 1:1 ratio or less than a 1:1 ratio. For example, each signal element can interface with (e.g., map to) greater than 1 cell in the set of cells 320. The set of cells 320 are preferably distributed between scaffolds in multiple layers (e.g., 2-10 layers, 10-20 layers, more than 20 layers, etc.), but can alternatively be distributed in a single layer. In an example, the number of cells between each pair of scaffolds (e.g., within a trench) can be between lo-imillion or any range or value therebetween (e.g., at least 100, at least ik; at least 5k; at least 10k, at least 20k, etc.). In an example, the total number of cells across the set of scaffolds 200 can be between looo-iomillion or any range or value therebetween (e.g., o.5million-6million). In an example, the cell seeding density can be between 100 cells per mm3 - 500,000 cells per mm3, or any range or value therebetween (at least 500 cells per mm3; at least 1000 cells per mm3; at least 5000 cells per mm3; at least 10,000 cells per mm3; at least 50,000 cells per mm3, at least 100,000 cells per mm3, etc.). Inan illustrative example, the cell seeding density can be similar to the cell density of native brain tissue.

[0072] The cell support 340 can function to: retain the bodies of the set of cells 320 within trenches in the set of scaffolds 200, support projections from the set of cells 320 into the native tissue, provide an interface between the set of cells 320 and the native tissue, diffuse molecules (e.g., nutrients) between native tissue and the set of cells 320, and / or perform other functions. In a specific example, the cell support 340 (e.g., gel) can be configured to support growth of axons and / or dendrites of the set of cells 320 out of the cell support 340. The cell support 340 is preferably a gel (e.g., a hydrogel), but can additionally or alternatively include any other suitable material. The cell support 340 material can include alginate, any biorthogonal hydrogel, and / or any other gel material. The cell support 340 is preferably fully transparent, but can alternatively be partially transparent (e.g., translucent) or non-transparent. The cell support 340 is preferably porous (e.g., microporous), but can alternatively be non- porous. The cell support 340 can optionally include growth factors, differentiation factors, small molecules, proteins, angiogenesis factors, extracellular matrix, nutrients, other chemicals, and / or other components. The cell support 340 can be positioned: between adjacent pairs of scaffolds in the set of scaffolds 200 (e.g., within the trenches), as a capping layer at the proximal end of the set of scaffolds 200, and / or otherwise positioned. The cell support 340 can optionally be adhered to the set of scaffolds 200 using: adhesive, polymerization, cross-linking, and / or any other bonding or other adhesion methods. However, any other cell support 340 can be used.

[0073] The set of cells 320 is preferably loaded into the set of scaffolds 200 (e.g., into the trenches) prior to implantation of the system 10, but can alternatively be loaded at any other time. In a specific example, the set of cells 320 are loaded into the set of scaffolds 200 with a liquid, unpolymerized cell support 340 (e.g., hydrogel), wherein the cell support 340 is polymerized (e.g., using calcium and / or any other polymerization factor) after loading. In a specific example, the set of scaffolds 200 can be capped with additional the cell support 340 (e.g., hydrogel).

[0074] In a first variant, neurons (e.g., post-differentiation) are loaded into the set of scaffolds 200 (e.g., into the trenches). In a specific example, the set of cells 320 can be loaded into the set of scaffolds 200 by performing one or more of: placing theset of scaffolds 200 in a loading liquid (e.g., the loading liquid occupies the trenches between scaffolds), placing the set of cells 320 in a loading fixture, centrifuging cells from the loading fixture into the trenches between the scaffolds (e.g., the cells adhere to the scaffolds), washing all or a portion of the set of scaffolds 200, and / or any other suitable steps.

[0075] In a second variant, stem cells and / or neural progenitor cells are first loaded into the set of scaffolds 200 and then differentiated. Differentiation can occur before or after implantation. In variants, differentiating after loading can reduce stress on the set of cells. In an example, the stem cells and / or neural progenitor cells can grow into (polymerized or unpolymerized) cell support 340 in the set of scaffolds 200. In a first specific example, differentiation factors can be added to the cell support 340 to induce or complete differentiation of the cells. In a second specific example, differentiation of the set of cells 320 can be induced or completed by implanting the system 10 in or on the brain. However, the set of cells 320 can be otherwise loaded.

[0076] However, the tissue interface 300 can be otherwise configured.

[0077] Implanting the system 10 in a user can optionally include one or more of: performing an incision on the skull, performing a craniectomy, resecting the dura, partially or fully ablating the pia mater (e.g., ablating using a laser; mechanically ablating using a blade; etc.), positioning the proximal subsystem 12 (including the proximal controller module 120, the set of scaffolds 200, and the tissue interface 300) on the brain surface (e.g., with pressure) such that the tissue interface 300 is in contact with the brain surface, folding the dura over all or a portion of the proximal subsystem 12, anchoring and / or sealing the proximal subsystem 12 to tissue, positioning the distal subsystem 14, anchoring and / or sealing the distal subsystem 14 to tissue, suturing skin around or over the distal subsystem 14, and / or any other suitable implantation steps.

[0078] The system 10 can optionally include a power source, which functions to power the controller 100 (e.g., the proximal controller module 120, the distal controller module 140, etc.), the set of scaffolds 200, and / or any other implanted system components. In a first variant, the power source is implanted (e.g., a battery implant). In a second variant, the power source (or a component of the power source) is remote to the implanted system components. For example, the power source caninclude a power transmitter component remote to the implanted system components (e.g., coupled to the external device 400, coupled to a separate device, etc.) as well as a power receiver component connected to one or more implanted system components (e.g., coupled to the controller 100). In specific examples, power receiver components can include: a magnetic coil, a magnetoresistive film, and / or other power receiver components that can interface with an external induction power source. In examples, the power source can provide power to the implanted system components via induction, IR, RF, and / or any other remote power method. However, the power source can be otherwise configured.

[0079] In an example, the power consumption of the implanted system components can be between lomW-ioomW or any range or value therebetween (e.g., 30mW-35mW, less than 50mW, etc.).

[0080] The system 10 can optionally include or be used with an external device 400, which can function to: determine the control instructions (e.g., excitation parameters), receive data (e.g., from the controller 100), process data, and / or perform any other functions. Examples of external devices that can be used include: a user device, a processing system (e.g., remote processing system), and / or any other suitable external device. The external device 400 can optionally include one or more of: communication elements configured to communicate with the controller 100 (e.g., with the distal controller module 140), processing system, sensors, power components, and / or any other suitable components. Information transmitted by the external device 400 can include: content (e.g., processed or unprocessed content), control instructions (e.g., calculated based on the content), and / or any other information. Information received by the external device 400 can include: data (e.g., cell states measured by the sensor elements), controller measurements (e.g., controller states measured by the controller sensor), and / or any other information.

[0081] However, the external device 400 can be otherwise configured.

[0082] In variants, the system 10 can include systems and methods as described in Appendix A, which is incorporated in its entirety by this reference. In variants, the system 10 can include systems and methods disclosed in: “A thin-film optogenetic visual prosthesis” (Knudsen EB, Zappitelli K, Brown J, Reeder J, Smith KS, Rostov M, Choi J, Rochford A, Slager N, Miura SK, Rodgers K, Reed A, Israeli YRL, Shiraga S,Seo KJ, Wolin C, Dawson P, Eltaeb M, Dasgupta A, Chong P, Charles S, Stewart JM, Silva RA, Kim T, Kong Y, Mardinly AR, Hodak M. 2023. bioRxiv doi: 10.1101 / 2023.01.31.526482), which is incorporated in its entirety by this reference.

[0083] However, the system 10 can be otherwise configured.5. Specific Examples.

[0084] A numbered list of specific examples of the technology described herein are provided below. A person of skill in the art will recognize that the scope of the technology is not limited to and / or by these specific examples.

[0085] Specific Example 1. A system configured to be implanted on a surface of a brain of a user, comprising: a set of genetically modified cells; a set of scaffolds, wherein the set of genetically modified cells are seeded between adjacent pairs of scaffolds in the set of scaffolds, wherein each scaffold in the set of scaffolds comprises: an array of recording electrodes configured to receive signals from the set of genetically modified cells; and an array of pLEDs configured to transmit light signals to the set of genetically modified cells based on control instructions; and a controller communicatively connected to the set of scaffolds, wherein the controller is configured to: receive data from the array of recording electrodes; and determine control instructions for the array of pLEDs.

[0086] Specific Example 2. The system of Specific Example 1, wherein the genetically modified cells comprise genetically modified neurons derived from pluripotent stem cells.

[0087] Specific Example 3. The system of any of Specific Examples 1 or 2, wherein the system further comprises a gel between adjacent pairs of scaffolds in the set of scaffolds, wherein the gel is configured to retain the genetically modified cells.

[0088] Specific Example 4. The system of Specific Example 3, wherein the gel is configured to support growth of axons of the genetically modified cells out of the gel.

[0089] Specific Example 5. The system of any of Specific Examples 1-4, further comprising an anchor configured to secure the set of scaffolds to the surface of the brain of the user.

[0090] Specific Example 6. The system of any of Specific Examples 1-5, wherein the controller comprises a proximal controller module and a distal controller module, wherein the proximal controller module is coupled to the set of scaffolds, and whereinthe distal controller module is anchored to the skull of the user, wherein the distal controller module is communicatively coupled to an external device.

[0091] Specific Example 7. The system of Specific Example 6, further comprising a flexible connector connecting the proximal controller module to the distal controller module.

[0092] Specific Example 8. The system of any of Specific Examples 1-7, wherein, for each scaffold in the set of scaffolds, the array of recording electrodes and the array of pLEDs are located on opposing faces of the scaffold.

[0093] Specific Example 9. The system of any of Specific Examples 1-8, wherein the genetically modified cells are transfected with a gene for a light-sensitive protein, wherein the genetically modified cells produce biochemical signals in response to receiving the light signals.

[0094] Specific Example 10. The system of any of Specific Examples 1-9, wherein at least 100 genetically modified cells are seeded between each pair of adjacent scaffolds.

[0095] Specific Example 11. The system of any of Specific Examples 1-10, wherein the set of scaffolds comprises at least 10 scaffolds.

[0096] Specific Example 12. A system configured to be implanted on a surface of a brain of a user, comprising: a first scaffold comprising an array of pLEDs on a first face of the first scaffold; a second scaffold comprising an array of recording electrodes on a first face of the second scaffold; a spacer separating the second scaffold from the first scaffold, the spacer positioned between the first face of the first scaffold and the first face of the second scaffold; a gel positioned between the array of pLEDs and the array of recording electrodes; and a set of genetically modified cells retained within the gel, wherein the array of recording electrodes is configured to receive signals from the set of genetically modified cells, wherein the array of pLEDs is configured to transmit light signals to the set of genetically modified cells.

[0097] Specific Example 13. The system of Specific Example 12, wherein the spacer comprises a first raised conductive element in contact with the first face of the first scaffold and second raised conductive element in contact with the first face of the second scaffold, the system further comprising a connector connecting the first raised conductive element to the second raised conductive element.

[0098] Specific Example 14. The system of any of Specific Examples 12-13, wherein the first scaffold further comprises a second array of recording electrodes on a second face of the first scaffold, wherein the second face of the first scaffold is opposite the first face of the first scaffold, and wherein the second scaffold further comprises a second array of pLEDs on a second face of the second scaffold, wherein the second face of the second scaffold is opposite the first face of the second scaffold.

[0099] Specific Example 15. The system of any of Specific Examples 12-14, wherein the set of genetically modified cells comprise axons extending out of the gel, wherein the axons interface with native neurons in the brain of the user.

[0100] Specific Example 16. The system of any of Specific Examples 12-15, wherein a thickness of the spacer is at least 100 pm.

[0101] Specific Example 17. The system of any of Specific Examples 12-16, wherein the genetically modified cells are transfected with a gene for a light-sensitive protein, wherein the genetically modified cells produce biochemical signals in response to receiving the light signals.

[0102] Specific Example 18. The system of Specific Example 17, wherein the set of cells comprise hypoimmune cells, wherein the set of cells are further transfected with a killswitch gene.

[0103] Specific Example 19. The system of any of Specific Examples 12-18, wherein the array of pLEDs comprises at least 100 pLEDs.

[0104] Specific Example 20. The system of any of Specific Examples 19, wherein the array of recording electrodes comprises at least 100 recording electrodes.

[0105] As used herein, "substantially" or other words of approximation (e.g., “about,” “approximately,” etc.) can be within a predetermined error threshold or tolerance of a metric, component, or other reference (e.g., within + / -o.ooi%, + / - 0.01%, + / -o.i%, + / -1%, + / -2%, + / -5%, + / -io%, + / -15%, + / -20%, + / -3O%, any range or value therein, of a reference).

[0106] All references cited herein are incorporated by reference in their entirety, except to the extent that the incorporated material is inconsistent with the express disclosure herein, in which case the language in this disclosure controls.

[0107] Different subsystems and / or modules discussed above can be operated and controlled by the same or different entities. In the latter variants, differentsubsystems can communicate via: APIs (e.g., using API requests and responses, API keys, etc.), requests, and / or other communication channels. Communications between systems can be encrypted (e.g., using symmetric or asymmetric keys), signed, and / or otherwise authenticated or authorized.

[0108] Alternative embodiments implement the above methods and / or processing modules in non-transitory computer-readable media, storing computer- readable instructions that, when executed by a processing system, cause the processing system to perform the method(s) discussed herein. The instructions can be executed by computer-executable components integrated with the computer-readable medium and / or processing system. The computer-readable medium may include any suitable computer readable media such as RAMs, ROMs, flash memory, EEPROMs, optical devices (CD or DVD), hard drives, floppy drives, non-transitory computer readable media, or any suitable device. The computer-executable component can include a computing system and / or processing system (e.g., including one or more collocated or distributed, remote or local processors) connected to the non-transitory computer-readable medium, such as CPUs, GPUs, TPUS, microprocessors, or ASICs, but the instructions can alternatively or additionally be executed by any suitable dedicated hardware device.

[0109] Embodiments of the system and / or method can include every combination and permutation of the various system components and the various method processes, wherein one or more instances of the method and / or processes described herein can be performed asynchronously (e.g., sequentially), contemporaneously (e.g., concurrently, in parallel, etc.), or in any other suitable order by and / or using one or more instances of the systems, elements, and / or entities described herein. Components and / or processes of the following system and / or method can be used with, in addition to, in lieu of, or otherwise integrated with all or a portion of the systems and / or methods disclosed in the applications mentioned above, each of which are incorporated in their entirety by this reference.

[0110] As a person skilled in the art will recognize from the previous detailed description and from the figures and claims, modifications and changes can be made to the preferred embodiments of the invention without departing from the scope of this invention defined in the following claims.APPENDIX AOptogenetic stimulation of a cortical biohybrid implant guides goal directed behaviorJennifer Brown, Kara Zappitelli, Paul Dawson, Eugene Yoon, Seton Shiraga, Yifan Kong, Max Hodak, Alan MardinlyAbstractBrain computer interfaces (BCIs) hold exciting therapeutic potential, but tissue damage caused by probe insertion limits channel count. Biohybrid devices, in which the cell-device interface is crafted in the laboratory, hold promise to address this limitation, but these devices have lacked a demonstration of their applicability for BCI. We developed a biohybrid approach to engraft optogenetically-enabled neurons on the cortical surface housed in a 2D-scaffold of circular microwells. The engrafted neurons survived, exhibited spontaneous activity, and integrated with the host brain several weeks after implantation. We then trained mice with biohybrid implants to perform an optical stimulation task and showed that they could effectively report optogenetic stimulation of their neural graft. This demonstration shows that a cortical biohybrid implant can be used to transmit information to the brain of the implanted animal.IntroductionBrain-computer interfaces have shown promise for numerous clinical applications1, but their efficacy is limited by the small number of communication channels with the brain23. The density of neural recording channels is constrained by a number of factors, including brain damage caused by insertion of neural probes into the parenchyma. Tissue destruction following probe insertion causes an inflammatory response characterized by hypoxic injury, mobilization of microglia around the insertion site, and formation of a glial scar. These reactions result in neuron loss in the vicinity of recording probes, as well as vascular and glial damage that limits the duration that chronic implants can successfully record from neurons4-9.Biohybrid neural interfaces are an emerging class of technology designed to circumvent these limitations by growing cells in a device in a laboratory and creating the device-host interface through a biological substrate10 11. Biohybrid devices are implanted into the target tissue in a manner allowing the implanted cells to engraft with host tissue and the embeddedelectrodes to record from and stimulate the engrafted cells. This approach holds potential advantages over traditional chronic neural probes; by allowing the engrafted cells to grow into the target tissue, probes can be arrayed at potentially far higher density than conventional probes. Neuron grafts are well known to integrate with the brain and acquire functional properties similar to the surrounding cortex, in some cases with therapeutic benefit12-17. However, neuron engraftment using microinjection has similar drawbacks to inserting neural probes: insertion of injection needles causes acute neuroinflammation and cell death early after engraftment18-21. Cortical biohybrid interfaces have been attempted using several different architectures, including seeding neural progenitor cells on silicon probes22 23, growing in projections from cells located distal to the cortex24-27, or by transplanting whole organoids and using an optical interface28-31. Biohybrid interfaces have also been implanted in the peripheral nervous system, including for regeneration of the neuromuscular junction32-34and peripheral nerve fibers3536. While biohybrid neural implants have exciting potential for BCI applications, they have so far lacked a functional proof-of-concept that these implants can transmit information to the brain, a necessary precondition for use as a brain interface. A biohybrid implant used for this purpose would have to overcome three challenging problems: 1) the implant must have an architecture that could allow accessing a very large number of neurons at high resolution. 2) The transplanted neurons must exhibit a high rate of survival after engraftment. 3) The grafted neurons must functionally integrate with the brain in a manner that allows the implanted organism to access and use information from the graft.To assess the suitability of this class of implants for BCIs, we designed a cortical biohybrid implant based on a planar array of microwells3738designed to hold one neuron per well. We implanted microwell scaffolds loaded with neurons on the surface of the cortex, and observed robust survival and graft integration within several weeks after implantation. To assess functional engraftment into the brain, we developed an optogenetic stimulation task39-41to test if mice could behaviorally report graft stimulation. Mice were able to perform this task to obtain rewards and sustain a positive bit rate. This results provides a proof-of-concept demonstration that a cortical biohybrid interface can provide input sufficient to drive goal-directed behavior, and opens a path for development of a class of high-bandwidth biohybrid neural interfaces.ResultsTo test the ability of a biohybrid implant to drive behavior, we designed an implant inspired by chronic mouse two-photon (2P) imaging, in which a piece of the skull is replaced by a glass coverslip42. We fabricated microwells designed to house single neurons and bondedthem to the bottom of the coverslip so that the neurons loaded in the wells could make contact with the brain, while also remaining optically accessible for light stimulation delivered through the coverslip. The microwell cell scaffolds were fabricated on fused silica wafers with SU-8 photoresist using photolithography. They were subsequently diced into 5x5mm dies, and then bonded to a glass coverslip with diameter 7 mm (Figure 1a). The microwells in the scaffold are packed in a hexagonal lattice at a 15 pm pitch. The interior of each microwell is a circle with a 10 pm diameter; the side walls are 2.5 pm thick and 8.6 pm long (Figure 1b). This configuration allowed dense packing of cells in the scaffolds (approximately 1.18 x 105microwells in an implanted surface of 25 mm2). The wells are 15 pm deep. To load neurons into these microwell scaffolds, we developed fixtures based on a PDMS stamp to create a seal around the coverslip. These fixtures contained a large reservoir to allow us to load the implant from a cell suspension pipetted on top of the scaffold, followed by capping and centrifugation under sterile conditions (Figure 1c).To obtain neurons for loading into our microwells, we performed embryonic dissections to obtain primary cortical neurons from C57 / B6J mice at E14.5-E15.543. This source of neurons had several advantages. Since these cells would be implanted back into other C57 / B6J mice, this allowed for an autologous cell graft and obviated the need for any immunomodulatory therapies at this stage of development. Furthermore, at this embryonic stage, few inhibitory neurons have migrated into the cortex, so the transplanted cell population should be primarily glutamatergic excitatory neurons4445. Indeed, we observed no teratomas or other overgrowth of the graft. To test our ability to load these embryonic primary neurons into our scaffolds, we performed dissections to obtain a cell suspension, and then used the fixtures to centrifuge cells into microwells. We allowed neurons loaded in microwells to recover in a CO2incubator overnight, and the array was evaluated with light microscopy the next day. Imaging without counterstains could not reliably determine whether individual wells contained cells (Supplemental Figure 1), so we fixed the arrays and used DAPI to determine how many wells contained a cell nucleus. These experiments determined that on average we filled 77±15% (mean ± s.d., n=23 microwell scaffolds, Figure 1d-e). With this loading density, the mean implants contained ~9 x 104neurons.To implant the cell-loaded scaffold in mice, we performed a craniectomy followed by duratomy, then placed the cell scaffold on the brain, sealed the implant, and added a head post (Figure 2a, see methods). 2P imaging of implanted mice showed no visible sign of structural deterioration of the scaffold even months after implantation. Scaffolds exhibited slight autofluorescence, but were optically transparent and overall compatible with 2P imaging of thebrain beneath the implants (Figure 2). We next transduced embryonic neurons with an adeno-associated virus (AAV) to allow visualization of the engrafted cells in scaffolds using in vivo 2P microscopy. After primary cell dissociation and scaffold loading, AAVs (AAV2 / 2-Syn-CheRiff-eGFP and / or AAV2 / 1-Syn-NES-jRGeco1a4647) were incubated with the loaded cell scaffold overnight, followed by surgical implantation. Just prior to implantation, the scaffolds were washed in PBS to avoid introducing viral particles to the brain. Control experiments in which empty scaffolds were loaded with virus in the absence of cells failed to observe any labeled cells in the brain, suggesting that this approach was sufficient to avoid virally transducing the brain (Supplemental Figure 2b).We next performed in vivo 2P imaging to assess graft survival and integration. Since cells did not express fluorescent proteins at the time they were surgically implanted, owing to the temporal delay in AAV expression after transduction48, we evaluated the success of the graft three weeks after surgery. Imaging three weeks after surgery, we found 50±24% (mean±s.d.) of microwells to be loaded with neurons expressing eGFP (n=13 implants, Figure 2b). Although this constituted a significant reduction in loaded wells relative to scaffolds loaded for 24 hours before implantation (Figure 2b, p<0.005, two-sided t-test), it still represents robust survival of the graft, as central nervous system graft survival rates are typically less than 25%18-20. 2P imaging of tissue below the microwells made clear that engrafted neurons extended extensive processes into the superficial layers of cortex (Figure 2d,f, Supplemental Figure 2a). Imaging of microwell grafts that had been incubated with virus and no cells or cells with no virus indicates that the fluorescence we observed in the superficial cortex was due to growth and integration of neurons expressing eGFP embedded in the implant (Supplementary Figure 2).To assess whether grafted neurons were active, in a small subset of experiments (n=3 mice), we transduced neurons with a calcium indicator jRGECOI a prior to graft implantation. 2P imaging of these scaffolds in awake mice head-fixed and running on a treadmill revealed spontaneous calcium events in these neurons three weeks after implantation (Figure 3a-b). Clustering of the pair-wise correlation coefficient between these calcium traces revealed multiple functional ensembles49, indicating that activity was not wholly correlated or uniform (Figure 3c). This data suggests that engrafted cells are able to mature and become active in the implant. We next assessed the degree of graft integration via histology. Unfortunately, dissecting the implant out of the skull without destroying the underlying cortex proved extremely difficult. Microwell arrays could not be left in place since they were adhered to a glass coverslip that prevented sectioning. When they were removed, they routinely fractured and pulled large chunks of cortical tissue out with them. Examination of explanted arrays indicated that cells remained inmicrowells even weeks after implantation. The presence of blood vessels in the tissue pulled from the brain during explant revealed tight mechanical coupling between the cortex and implanted cells in microwell arrays (Figure 3d). Although the damage to cortex caused by explant made systematic histology impossible, in some animals we were able to obtain some sections near the implant site and were able to confirm our observations via 2P microscopy of robust integration in the superficial cortex. Additionally, we were able to observe axons throughout the layer of the cerebral cortex, suggesting that neurons implanted above layer 1 are able to extend axons deep into the brain (Figure 3e).Since our engrafted neurons survive and exhibit spontaneous activity, we next sought to determine if the graft was functionally integrated with the host brain. To accomplish this, we elected to design a behavioral task in which the mouse would be required to use the activity of the biohybrid neurons to obtain rewards. In this task, an animal initiates a trial by activating a center nose-poke port, and then must report whether optogenetic stimulation occurred by activating a left or right port to obtain a water reward3940(Figure 4a-b, see methods). To validate the task, we first trained two control cohorts. Positive control animals were injected with an AAV encoding an excitatory opsin in primary somatosensory cortex, while negative control animals were injected with an AAV expressing a calcium indicator instead, which should not alter neural activity in response to light. Four weeks later, we installed a cranial window, screened for expression, and installed a fiber optic ferrule attached to an LED. We observed that 86% (6 of 7) positive control mice injected with AAV1-eSYN-hChR2(H134R)-eGFP (n=5) or AAV2 / 2-Syn-CheRiff-eGFP (n=2) reached criterion performance (d’ > 1.25) within the three week training window (9.5±3.5 days to criterion mean±s.d.). In contrast, none of the mice in the negative control cohort injected with AAV2 / 1-Syn-NES-jRGECO1a (n=4) reached criterion during the three weeks of training. This confirms that optogenetic activation of neurons using a fiber-coupled LED was required to successfully complete the task, and that animals were unable to use the presence of light, sound, heat, or other external stimuli as cues to complete the task (Figure 4c).We then trained a cohort of animals implanted with our biohybrid implant to complete the task (n=9). These mice were implanted with neurons incubated overnight with AAV2 / 2-Syn-CheRiff-eGFP. Three to four weeks later, animals were screened for expression and then implanted with a ferrule and started on behavior training. Five of nine mice in this cohort passed criterion performance in the three weeks training period (Figure 4c-e, not statistically significant relative to the positive control group, Fisher Exact Test p=0.3). For animals that did complete their task, time to criterion performance was identical to the positivecontrol group (9.6±4 days to criterion mean±s.d., Supplementary Figure 3, p=0.95, t-test). To control for the possibility that residual virus in the microwells transduced neurons in the brain, we trained an additional control group where we implanted microwell scaffolds that had been incubated with AAV2 / 2-Syn-CheRiff-eGFP, but not loaded with cells. We did not observe fluorescence in these mice via 2P microscopy (Supplementary Figure 2b), and none (0 of 4 mice) of the animals in this group achieved criterion performance. This suggests that if any residual AAV did transduce cells in the host brain, it was not sufficient to drive detection of the light stimulus. Additionally, no (0 of 4) mice that received biohybrid implants without AAVs achieved criterion performance, indicating that engraftment alone is not sufficient to allow mice to detect optical stimuli (Figure 4c-e). Since we imposed a trial structure that allowed a variable response time after stimulus, we were able to retroactively assign a bit-rate to each mouse during active trial time. Remarkably, biohybrid mice achieved a positive bit rate similar to positive controls and sustain bit rates as high at 0.7 bits I seconds (Supplementary Figure 3, best session performance for biohybrid rate averaged 0.25±0.24 bits per second mean±s.d., n = 9). Overall, this data supports the hypothesis animals implanted with biohybrid devices above cortical layer 1 have access to information originating from the activity of the graft, and are able to use that information to guide goal-directed behavior.DiscussionThis study provides a proof-of-concept demonstration that an optogenetic biohybrid neural implant positioned on the cortical surface can be activated by light and used for goal directed behavior. The implant consists of tens of thousands of neurons positioned above the cortex in microwell scaffolds. Since the number of surviving engrafted cells is roughly equivalent to the number of neurons in a cubic millimeter of mouse cortex50, it is perhaps not surprising that the animal is able to detect stimulation of the graft, as the stimulus could potentially generate hundreds of thousands of action potentials. Indeed, detection of optogenetic stimulation of the brain has been reported to require as few as several hundred neurons39, so our experiments do not provide a clear readout of the degree of graft integration with the cortex. While it is likely that the graft forms chemical synapses with the host brain that are used to read out the state of the engrafted cells, we cannot rule out the possibility that the mouse uses other cellular mechanisms to complete the task, such as detection of bulk glutamate release. Regardless, since neither implantation of the empty virus-washed scaffold nor opsin-negative cells is sufficient to complete the task, the animals clearly have some way to use the graft to detectactivation of the LED. Mice can learn to act on optical stimulation of many brain areas40, and similarly to optogenetic stimulation of various brain areas, optogenetic activation of engrafted cells likely provides a perturbation to on-going activity that the animal is able to monitor. Not all of the engrafted mice learned to perform the task. Opsin expression and survival of the graft cannot explain this result, as animals were pre-screened for expression before entering training, and the number of observed cells was not correlated with task performance (Supplementary Figure 4a). However, the engrafted animals that did learn the task performed just as well as the positive control group and learned just as quickly (Supplementary Figure 4b).Surprisingly, our grafts exhibited very high rates of neuronal survival (~50%). Engraftment of neurons via microinjection results in only a small fraction of surviving cells across a wide range of neuronal cell types and brain regions18-21 51,52. Apoptosis after engraftment53is thought to be driven by hypoxia and immune insult associated with disruption of the blood brain barrier21, and occurs in the first few days after engraftment1852. By placing our implant on the cortical surface, we minimize the disruption of the vascular network, and therefore avoid the major mechanism of cell death after transplantation. Additionally since our graft is one cell-layer thick and spread evenly across the cortex, we avoid the deleterious microenvironment thought to be present in the center of a bolus of injected cells. Since quantification of microwell loading required fixing cells, we were unable to measure the number of microwells loaded with neurons pre- and post-implantation to quantify the number that survived transplantation. Similarly, use of a viral transgene to visualize our graft neurons meant that we were not able to measure graft survival early within the first weeks after surgery. Some apparently empty wells likely have healthy neurons that were not transduced by the virus, or may have been occupied by cells unable to express proteins downstream of the synapsin promoter. We note that the use of allogeneic embryonic primary neurons in this study may be a key factor governing graft survival, and is not an option for translational approaches.Although placing the biohybrid device on the cortical surface provided key advantages for graft survival and future integration with devices, it caused unique challenges for graft integration. Specifically, in order to integrate with the brain, neurons needed to extend processes through the pia mater, which specifically functions to prevent infiltration into the brain54. Indeed, most grafted cell processes were in the superficial cortex, although sparse axons were detected throughout the cortex (Figure 3e). Occasionally, labeled cell bodies were detectable in the superficial cortical layers. We cannot tell if these are engrafted neurons that migrated out of their wells or if they are the result of residual AAV sparsely transducing the cortex. Our behavioral control cohorts suggest that even if these labeled cells are the result ofresidual virus, they are not sufficiently numerous to allow the animal to perform the detection task. The presence of numerous neurons in microwells months after engraftment suggests that migration is rare, but we cannot categorically rule out some degree of migration out of the wells and into the cortex.The ability to learn to detect graft stimulation implies that an animal has access to the information conveyed by the engrafted neurons’ activation and can use it to guide a goal directed behavior, in this case to obtain water rewards. Thus, this report demonstrates the first example of a biohybrid implant’s use in a brain computer interface task. In this report, we elected to leave the top of the implant optically transparent so as to allow live imaging of neurons in the graft, and behavior was driven by commercial fiber-coupled LEDs. However, we and others have fabricated high density pLED displays55at similar pitch to the microwell scaffolds. Future versions of a biohybrid implant could allow pixels to be aligned to microwells to allow stimulation at near single-cell resolution. Although numerous challenges exist towards translating biohybrid neural interfaces for extremely high bandwidth BCIs, this study represents a proof-of-concept demonstration that such devices can in principle be used to guide goal directed behavior.MethodsAnimalsC57 / B6J male and female mice obtained from Charles River were used in this study. Mice were maintained on a 12 hour light-dark cycle. All animal procedures were carried out with the approval of Science Corporation's institutional animal care and use committee.Fabrication of microwell implantsMicrowells were fabricated on 100 mm fused silica wafers using photolithography. In preparation for lithography, the wafers were cleaned using a spin rinse dryer, dehydrated for 15 minutes on a 200°C hotplate, and exposed to O2plasma in a Technics RIE for 2 minutes at 50W and 300mTorr. To improve microwell adhesion, a thin layer of SU-82001 (Kayaku Advanced Materials) was spun at 3000 rpm and UV flood exposed to anchor the subsequent thicker layer. To achieve a microwell height of 15 pm, SU-8 2010 photoresist (Kayaku) was spun at 3000 rpm for 30 seconds, soft-baked at 95°C for 3.5 minutes on a hotplate, and exposed with a Karl SussMAB6 mask aligner using contact lithography and a total UV dose of 294mJ / cm2. Exposure was done utilizing a long pass filter to filter wavelengths below 360 nm. We completed a post exposure bake of 4 minutes at 95 °C and developed for 3.5 minutes in SU-8 Developer (Kayaku). To improve biocompatibility56, a UV flood-exposure was used with a total UV energy of -1.86 J. Finally, the wafers were hard baked by holding at 95°C for 5 minutes and then ramping to 200°C for 2 minutes on a hotplate. Wafers were then coated in SPR220-7 photoresist as protection for dicing and were diced into 5x5mm die using a Disco DAD3240.Individual microwell die were then mounted to 7mm diameter glass coverslips using a small drop of Epotek 301 epoxy and cured in an oven at 65°C for 2 hours.Loading fixtures were fabricated using 3D printing and CO2laser cutting techniques. All components not in contact with the biological solution were FFF (Fused Filament Fabrication) printed (Raise3d ProPlus printer, commercial PLA+ filament [ANYCUBIC PLA+ Gray or equivalent]) . Nominal hot-end temperature was 205°C, and the bed was heated to 50°C. The bed interface was treated with a film of PVA (Elmer’s All Purpose Glue Stick) to aid in part-to-bed adhesion. After printing, all FFF parts were scrubbed and soaked in deionized water for a minimum of one hour. Gaskets were made from FDA 177.2600 approved silicone sheets. A CO2laser cutter (Full Spectrum Muse) was used to cut the gasket shape. After laser cutting, all gaskets were soaked in deionized water for a minimum of one hour. The well was fabricated using a photopolymerization printer (FormLabs 3B, BioMed Clear resin) according to manufacturer’s recommendations, followed by a one hour soak in deionized water..Preparation of loaded biohybrid scaffoldMicrowell arrays were sterilized in 70% ethanol and left to dry under UV light exposure overnight in a biosafety cabinet. Prior to loading neurons, microwell arrays were loaded with 0.1 mg / mL poly-d-Lysine (Gibco) and 0.1 mg / mL Laminin (Millipore Sigma, L2020) in DI water for at least one hour. Following incubation, arrays were washed three times with sterile water and placed in a CO2incubator in a sterile petri dish.Embryonic cortical neurons were prepared essentially as described43. Briefly, pregnant C57 / B6 mice at embryonic day E14.5-E16.5 were sacrificed via CO2euthanasia, sprayed down with 70% ethanol then transferred to a sterile dissection hood. The uterine sac was transferred to ice cold dissection media (HBSS supplemented with 100 mM MgCI2*6 H2O, 100 mM HEPES,and 15 mM Kyneuric Acid adjusted to pH 7.2). Embryos were removed from the uterine sack, and the brains were dissected out under a stereoscope using a pair of fine forceps. The meninges were removed, and the cortices collected in ice cold dissection media. After all cortices were collected, the tissue is incubated in papain solution (dissection media supplemented with 2.5 mM L-cysteine and 100 units of Papain (Millipore Sigma) in a CO2incubator for 5-7 minutes. The tissue was carefully washed three times with prewarmed light inhibitor solution (dissection media with 6 mg / mL bovine serum albumin and 6 mg / mL Trypsin inhibitor followed by three washes with heavy inhibitor solution (dissection media with 60 mg / mL bovine serum albumin and 60 mg / mL Trypsin inhibitor). Following the heaving inhibitor wash, the tissue is washed once with prewarmed Neurobasal medium and then triturated into single cell suspension in Neurobasal+ media (neurobasal supplemented with 1% Penicillin-Streptomycin, 1% Glutamax, and 2% B27 supplement. Cell concentration was counted and adjusted to 1e6 neurons I mL and kept on ice.Microwell arrays were removed from the incubator, and assembled in a loading fixture in the biosafety cabinet using sterile technique. The neuron suspension was transferred into a loading fixture, covered, and centrifuged at 300 ref for 5 minutes at room temperature. The loaded microwell array was incubated in a CO2incubator for two hours, and then removed from the fixture in a biosafety cabinet two hours later and placed into a standard cell culture dish with fresh prewarmed neurobasal+. In some experiments, AAVs (see below) were added to the media for an overnight incubation prior to surgical implantation the next day.Viral InjectionsFor all surgical procedures, mice were anesthetized with isoflurane (2%), and administered 2 mg / kg of dexamethasone as an anti-inflammatory and 0.05 mg / kg Ethiqa as an analgesic Animals were maintained on isoflurane on a heating pad and head fixed in a stereotactic apparatus (Kopf). After sterilizing the incision site, the skin was opened, and a small burr hole was drilled over S1 (ML 2mm, AP +2mm from Lambda) using a 0.24 mm drill bit (Busch). 750-1000 nL of virus was injected using a micro syringe pump (Micro4) and a wiretrol II glass pipette (Drummond) at a rate of 25 nL / second 3-4 depths 1500-500 pm below the brain surface. After the injection was complete, the needle was held in place for several minutes before suturing the scalp closed. Viruses used: AAV2 / 2-Syn-CheRiff-eGFPP (Biohippo #BHV12400513, titer: 2.37e12 vg / mL), AAV1-eSYN-hChR2(H134R)-eGFP (Biohippo #BHV12400006, titer: 1.6e13 vg / mL), AAV2 / 1-Syn-NES-jRGeco1a (Biohippo #BHV12401534,titer: 5.85e12 vg / mL). 25-30 days after virus injection, mice were anesthetized, and a cranial window and headplate was installed over the injection site. The scalp was removed, and the fascia retracted. Following application of Vetbond (3M) to the skull surface, a custom stainless steel headplate was fixed to the skull with two dental cements: Metabond (C&B) followed by UV-cure acrylic (Flow-IT ALC). After the dental cement dried, a 3 mm diameter craniotomy over the left primary somatosensory cortex was drilled, and residual bleeding stopped with repeated wet-dry cycles using sterile hypotonic saline, gauze, and Gelfoam (Pfizer). A window plug consisting of two 3mm diameter coverslips glued to the bottom of a single 5mm diameter coverslip (using Norland Optical Adhesive #71) was placed over the craniotomy and sealed permanently using Metabond and Flow-IT. Animals were allowed to recover in a heated recovery cage before being returned to their home cage. Approximately three days after surgery, animals were habituated to head fixation under a freely moving circular treadmill and screened for virus expression under an upright epifluorescence microscope (Thorlabs Cerna, 10x mitutoyo objective 0.28 NA).Microwell ImplantsMice were anesthetized with isoflurane (2%), and administered 2 mg / kg of dexamethasone as an anti-inflammatory and 0.05 mg / kg Ethiqa as an analgesic. The scalp was removed, the fascia retracted, and the skull lightly etched. A 5 mm diameter circular craniotomy was drilled over primary somatosensory cortex, and residual bleeding stopped with gelfoam (Pfizer). A duratomy was performed using a von graefe knife, and the dura carefully retracted using fine forceps. Residual bleeding was again stopped by gelfoam. Microwell implants loaded with cells the day before surgery were picked up with forceps, and gently washed three times with sterile neurobasal media or PBS before being inverted and placed into the craniotomy. Downward pressure was applied using a stereotaxic attachment, and the interface between the 7 mm outer coverslip and the skull filled with Metabond (C&B). After the metabond dried, a custom stainless steel headplate was fixed to the skull using Metabond (C&B) and Flow-IT ATC. Animals were allowed to recover in a heated recovery cage before being returned to their home cage. Animals were evaluated using a 2P microscope approximately three weeks after surgery.Ferrule implantationFollowing screening for expression and / or engraftment, mice were anesthetized with isoflurane (2%) and mounted in a stereotaxic frame. A 400 pm fiber optic cannula (Thorlabs CFMLC14L02, 0.39 NA) was secured to the top of the glass cranial window using several drops of UV-cured optical acrylic (Norland 71). The implant was then covered in Flow-IT ALC, cured, and then covered with metabond (C&B) mixed with black oxide to make it optically transparent. Animals were allowed to recover from anesthesia in a heated recovery cage before being returned to their home cage, and they began water restriction for behavior several days later.2P imagingFor 2P imaging, mice were head-fixed on a freely spinning running wheel under a Nixon 16x-magnification water immersion objective (N16XLWD-PF) and imaged with a custom built 2P resonant scanning microscope within a darkened box. The running wheel setup was mounted on a Hexapod (Physik Instrumente) which allowed the plane of the implant to be roughly matched to the microscope imaging plane. Structural imaging was performed using Scanlmage5 (Vidrio), and z-stacks were acquired by moving the objective using a piezo actuator (PFM450E, Thorlabs). Imaging was performed at an average power of 100 mW at 930 nm using a Coherent Discovery femtosecond laser. For functional jRGECOIa imaging, single plane imaging was performed at 30 Hz using a wavelength of 1040 nm at an average power of ~50 mW. Images were analyzed without motion correction, since the imaging plane was stable and the hexagonal lattice structure of the microwells, which were always visible due to faint fluorescence, disrupted these operations and made the resulting motion corrected stacks far worse than the uncorrected stacks. Calcium sources were identified morphologically using CellPose57on an average image. Average calcium intensity values were converted to dF / F by baselining using the 10th percentile of intensity values of each calcium source.HistologyAnimals were deeply anesthetized using Ketamine and transcardially perfused with cold PBS followed by 4% paraformaldehyde. After perfusion, the head was removed and we attempted to remove the metabond holding the implants in place using a dental drill. Implants were removed using forceps, but we often observed strong adhesion between the implant and the brain. This adhesion resulted in numerous tissue rips and tears on the implant's surface, including numerous instances in which large chunks of brain under the cranial window were pulled outwhen the window was removed. After removing the implant, the rest of the brain was dissected out of the skull, and then post-fixed for 24 hours, followed by cryoprotection in sucrose gradients. After the brain sank in 30% sucrose it was snap frozen in OCT using a isopentane liquid nitrogen bath. Samples were left in the -20C freezer overnight. Using a Leica cryostat, 50pm cryosection retinal slices were made and mounted onto Fisherbrand™ Superfrost™ Plus Microscope Slides (Fisher Scientific Cat No. 12-550-15). Samples on slides were then blocked in a solution containing 10% donkey serum (DS), 0.1 M Glycine, 0.3% Triton X-100 in PBS for 1 hour. Primary antibodies were diluted in 1% DS in PBS at room temperature overnight. Samples were washed in PBS three times for 5 minutes, then incubated for 2 hours at room temperature with secondary antibodies diluted in PBS. Samples were washed in PBS three times for 5 minutes, then 4',6-Diamidino-2-Phenylindole, Dihydrochloride (DAPI) (D1306 Thermofisher Scientific) (1:1000) was added to all the samples to stain cell nuclei and incubated at room temperature for 5 minutes. All samples were washed again with PBS for 5 minutes. Fluorescent images were captured using a ZEISS LSM 980 confocal microscope. Micrographs were captured using the Plan-Apochromat 10x / 0.45 air objective (Carl Zeiss Microscopy, Jena, Germany). ZEN Blue 3.5 software was used to perform stitching of tilescans and maximum intensity projections of Z-stacks.Optical Stimulation TaskAfter surgery, mice were placed on a high fat diet for at least one week. After baseline weight was established by measuring daily for three consecutive days, animals were restricted to 1 mL of water daily for 5-9 days before beginning behavior training. Throughout training, mice were weighed daily and given supplemental water if their weight fell below baseline for 1 day, and were given access to free water if their weight didn’t recover to at least 80% of baseline weight after supplemental water.The optical microstimulation task was conducted in a 9 x 9 inch box in which the animal was allowed to freely move. On one side of the box, three nose poke ports were installed. Each port contains an IR beam break. The center point contains masking light LEDs and the left and right ports contain a lick spout controlled by two solenoid valves (Nresearch) connected to a water reservoir calibrated to set each reward at 50 uL of water. Experimental control was conducted with an microcontroller (Arduino Due) mated to a custom circuit board. Data was streamed via serial port to a PC that allowed the experimenter to control the task via custom software. Prior to a session, the cover on animals’ implanted optical fiber ferrule was removed,and a 400 m optical fiber was inserted into the ferrule. The fiber was commutated by a rotary joint (Thorlabs RJPSF2) and attached to a Thorlabs LED (470 nm) controlled by a Thorlabs LEDD1 B driver. Optical power throughput was measured using a power meter before surgical implantation, and behavior was run at levels that corresponded to 5 mW of optical power out of the ferrule before surgical implantation. In some high performing animals, we reduced the power to 1 mW and they were still able to perform the task, but we did not systematically test the effect of optical power on performance.Mice were introduced to the task in three phases. In phase 1 , the animals learned to navigate the box with an optical fiber attached to the head and to collect water rewards from both the left and right port. In this phase, triggering the center nose port had no effect, but triggered the left or right nose ports resulted in an immediate water reward, with a variable 1-5 second cooldown period before the next reward was available. Animals are advanced to phase 2 as soon as they obtain 15 rewards from both ports in 10 minutes, demonstrating that they have the ability to trigger the IR beam break and obtain water.In phase 2, animals learned to initiate a trial in the center port and collect reward on the left or right ports. In this phase, rewards were only available on the left and right ports after the center beam had been broken. Since the mouse was now able to associate the noise from the solenoid valve with the availability of reward, each center poke triggered an immediate reward available on either the left or right port (changed randomly each trial). If the animal poked the correct port within 10 seconds, additional reward was given from that port. After 20 successful trials on each side in a session, animals were advanced to phase 3.Phase 3 is the complete task. Mice initiated trials by poking the center port, which triggered the immediate activation of masking lights (also activated in earlier phases) as well as optogenetic stimulation on some trials (10 pulses at 10 ms, 20 Hz). Reward was available on the left side for optogenetic stimulation or the right side for no stimulation. To avoid rewarding a strategy where a mouse could simply go to one side for consistent reward on 50% of trials, trial type was pseudorandomized to adjust for bias every 30 trials to encourage mice to engage with both the left and right ports. Trials were scored as HIT trials if it successfully reported optogenetic stimulation, MISS if it failed to select the correct report after stimulation, FALSE ALARM (FA) if it selected the stimulation port after a catch trial, or Correct Reject (CR) if it selected the right port during a catch trial.For each behavioral session, which lasted 30 minutes or -100 trials, we calculated the discriminability index d’. We compensated for trial types that didn’t occur by adjusting zeros by resetting trial types with zero events to 1 - 1 / 2N, where N = number of trials. We then calculatedhit rate (HR; hits / hits + misses) and false alarm rate (FAR; FA / FA + CR). We z-scored these using the inverse cumulative distribution function (CDF) of the standard normal distribution, and subtracting to find d -Z(HR)-Z(FAR). Criterion performance was achieving d’ > 1.25 for two consecutive trials, after which training was stopped. All training was halted after 21 sessions, or 8 sessions longer than the slowest-learning animal in the positive control group, and these animals were deemed to have not learned the tasks. Training occurred 5 or 6 days a week, with a 1 mL of free water administered on break days. Bit rate was calculated on a per-session basis using Wolpaw’s definition58for N=2 choices, bit_rate = (p + (1 - p) * np. Iog2( 1 - p)) / T where p is average success rate and average trial duration.Figures and Figure LegendsFigure 1: Fabrication and loading of microwell cell scaffolds a. Light micrograph of a microwell scaffold bonded onto a glass coverslip (scale bar is 100 pm) b. Electron micrograph of a cross-sectioned empty microwell cell scaffold taken at 45° tilt, sputtered with gold for imaging. Scale bar is 10 pm c. i) exploded view of the microwell loading fixture with labeled parts, ii) rendering of assembled loading fixture d. Low zoom confocal micrograph of DAPI-stained loaded microwell scaffold taken 24 hours after loading. Scale bar is 100 pm. e. Low zoom confocal micrograph of DAPI-stained loaded microwell scaffold taken 24 hours after loading. Scale bar is 25 pm. Asterixis indicate unloaded wells. f. Histogram showing the fraction of wells in individual scaffolds that were occupied by DAPI-positive staining 24 hours after loading (n=23 scaffolds).Figure 2: Robust survival and integration via cortical microwell scaffolds a. Cartoon demonstrating experimental setup wherein mouse primary neurons are labeled with an AAV expressing eGFP and loaded into microwell scaffolds mounted on coverslips which are subsequently implanted onto the surface of the mouse brain. b. Histograms showing the fraction of wells in individual scaffolds that were occupied by DAPI-positive staining 24 hours after loading (blue, n=25 scaffolds) and the fraction of wells containing eGFP-labeled neurons via 2P microscopy approximately 21 days after surgery (orange, n = 13 implanted scaffolds). The distributions are significantly different (p = 0.0005, two sided t-test) c. Single 2P image from a representative mouse loaded with a microwell implant approximately three weeks after surgery. The green channel shows virus-labeled eGFP positive engrafted cells, and the red channels collects autofluorescence from the implanted scaffold. The image plane is centered approximately on the center of the 20 pm cell scaffold. Scale bar is 100 pm. d. Single 2P image from the same mouse implanted in c, but here the image plane is centered approximately 50 pm below the cell scaffold above. Scale bar is 100 pm. e. As in c, 2P image centered on microwell plane, but at higher magnification and a different representative mouse three weeks after surgery. Scale bar is 25 pm. f. As in d, 2P image taken approximately 50 pm below the scaffold shown in e. Scale bar is 25 pm.<JF;F CorrelationFigure 3: Cells in microwell scaffolds mature and integrate a. Average image created from 2P functional imaging of implanted cells in microwell scaffold expressing jRGECOIa. Scale bar is 25 pm. b. Representative df / f traces from 30Hz 2P imaging of jRGECOIa-expressing cells loaded in a microwell scaffold. Figure shows spontaneous activity from a head fixed mouse free to run on a wheel. Scale bar is 10 seconds. Colorbar shows df / f values from 0 to 1. c. Representative dendrogram showing pairwise correlation coefficients from spontaneous activity of cells implanted in microwell scaffolds. Colorbar shows correlation coefficients from 0 to 1 . d. Confocal micrograph (single plane) of representative explanted microwell scaffolds months after implantation. Cells in this experiment were transduced by AAV expressingCheriff-eGFP (green) and the scaffold's intrinsic fluorescence dominates the blue channels. Explanting microwells was usually very destructive; in this example a region of the microwell scaffold outlined in the white dashed line fractured, but the tissue beneath was sufficiently adhered to the scaffold that it pulled up from the brain. The presence of blood vessels (white arrows) indicates that the brain was tightly adhered to the graft. The scale bar is 50 pm. The inset below shows loaded cells still in their microwells (white arrows), scale bar 10 pm. e. Confocal micrograph (maximal intensity projection) of representative region of cortex beneath and explanted microwell scaffold removed months after implantation. Cells in this experiment were transduced by AAV expressing Cheriff-eGFP (green) nuclei are labeled with DAPI (blue). Dense processes are observed in the superficial cortex, but putative axons are visible deep in the cortex (white arrows; inset). Scale bars are 50 pm and 25 pm in the inset. Asterix indicates the cell body in the cortex.Figure 4: Optical stimulation of the biohybrid implant guides goal directed behavior. a. Cartoon illustrating task in which mice implanted with optogenetic biohybrid microwell implants learn to detect stimulation of the grafted cells to gain rewards. Briefly, trials are initiated by breaking an IR beam at the center nose poke, which triggers masking lights on every trial. Rewards are administered after breaking the beam on the left port on trialswhere there is no stimulation and on the right port on trials in which there is optogenetic stimulation of the implant. b. Representative sequence of four trials, all correct, in which the mouse alternates between the right and left ports on stimulation and catch trials to obtain water rewards. c. Plots showing training day (x-axis) versus discriminability index d’ on the y-axis. On each plot a dashed black line indicates criterion performance. Each colored line is the score from an individual animal on a given training day. Subplot are positive control (AAV-excitatory opsin injected into cortex, n=7, red), negative control (jRGecola AAV injected into cortex, n=4, black), biohybrid implants (engrafted neurons on microwell scaffolds with AAV-Cheriff, n=9, blue), cells only (microwell scaffolds loaded with cells and no AAV, n=4, green), or virus only (microwell scaffolds incubated with AAV-Cheriff, n=4, magenta). Note that in each case animals were trained for 21 sessions or until they achieved criterion performance on two consecutive days. d. Box and swarm plots showing the discriminability index for the best session for each animal in each condition. Each dot corresponds to a single animal, and the boxplots show mean and interquartile range (p < 0.005, one-way ANOVA). e. Stackplot showing the total number of animals in each group that passed criterion performance (orange) or failed to pass criterion performance (blue) during the21 -session training window (p<0.005, Chi-Squared test).Supplementary FiguresSupplementary Figure 1 : Live imaging of loaded microwell scaffolds pre-implantation a. Confocal micrograph showing mouse embryonic neurons (gray, transmitted PMT) in SU-8 microwell scaffolds (green, scale bar 100 pm). b. Higher zoom confocal micrograph showing mouse embryonic neurons (gray, transmittedPMT) in SU-8 microwell scaffolds (green, scale bar 25 pm).Supplementary Figure 2: 2P Imaging of control microwell conditions a. Maximal intensity projection of a 2P image stack from a representative mouse implanted with a microwell scaffold (red) containing neurons incubated with an AAV-Cheriff-eGFP virus for 24 hours before implantation and then washed pre-surgery. Image is taken three weeks after surgery. Scale bar is 100 pm. b. Maximal intensity projection of a 2P image stack from a representative mouse implanted with a microwell scaffold (red) without cells incubated with an AAV-Cheriff-eGFP virus for24 hours before implantation and then washed pre-surgery. Image is taken three weeks after surgery. Scale bar is 100 pm. c. Maximal intensity projection of a 2P image stack from a representative mouse implanted with a microwell scaffold (red) loaded with cells without any virus. Image is taken three weeks after surgery. Scale bar is 100 pm.Supplementary Figure 3: Positive bit rates using a biohybrid BCI a. Plots showing training day (x-axis) versus mean bit-rate (bits per second) on the y-axis. Each colored line is the bit rate from an individual animal on a given training day. Subplot are positive control (AAV-excitatory opsin injected into cortex, n=7, red), negative control QRGECOIa AAV injected into cortex, n=4, black), biohybrid implants (engrafted neurons on microwell scaffolds with AAV-Cheriff, n=9, blue), cells only (microwell scaffolds loaded with cells and no AAV, n=4, green), or virus only (microwell scaffolds incubated with AAV-Cheriff, n=4, magenta).b. Box and swarm plots showing the bit rate for the best session for each animal in each condition. Each dot corresponds to a single animal, and the boxplots show mean and interquartile range. aSupplementary Figure 4: Time to learn optical stimulation task a. Box and swarm plots showing the fraction of loaded microwells in the biohybrid animals that learned or did not learn the optical microstimulation task, each red dot corresponds to one animal (P>0.9, two-sided t-test).b. Box and swarm plots showing the number of training sessions needed to achieve criterion performance. Only animals that achieved criterion performance are shown, each red dot corresponds to one animal (P>0.95, two-sided t-test).Works Cited1. Patrick-Krueger, K. M. The state of clinical trials of implantable brain-computer interfaces.2. Willett, F. R. et al. A high-performance speech neuroprosthesis. Nature 620, 1031-1036 (2023).3. Chen, X., Wang, F., Fernandez, E. & Roelfsema, P. R. Shape perception via a high-channel-count neuroprosthesis in monkey visual cortex. Science 370, 1191-1196 (2020).4. Ereifej, E. S. et al. Implantation of Neural Probes in the Brain Elicits Oxidative Stress. Front. Bioeng. Biotechnol. 6, 9 (2018).5. Polikov, V. S., Tresco, P. A. & Reichert, W. M. Response of brain tissue to chronically implanted neural electrodes. J. Neurosci. Methods 148, 1-18 (2005).6. Chen, X. et al. Chronic stability of a neuroprosthesis comprising multiple adjacent Utah arrays in monkeys. J. Neural Eng. 20, 036039 (2023).7. Kumosa, L. S. & Schouenborg, J. Profound alterations in brain tissue linked to hypoxic episode after device implantation. Biomaterials 278, 121143 (2021).8. Kozai, T. D. Y, Jaquins-Gerstl, A. S., Vazquez, A. L., Michael, A. C. & Cui, X. T. Brain Tissue Responses to Neural Implants Impact Signal Sensitivity and Intervention Strategies. ACS Chem. Neurosci. 6, 48-67 (2015).9. Gulino, M., Kim, D., Pane, S., Santos, S. D. & Pego, A. P. Tissue Response to Neural Implants: The Use of Model Systems Toward New Design Solutions of Implantable Microelectrodes. Front. Neurosci. 13, 689 (2019).10. Rochford, A. E., Carnicer-Lombarte, A., Curto, V. F., Malliaras, G. G. & Barone, D. G. When Bio Meets Technology: Biohybrid Neural Interfaces. Adv. Mater. 32, 1903182 (2020).11. Boulingre, M., Portillo-Lara, R. & Green, R. A. Biohybrid neural interfaces: improving the biological integration of neural implants. Chem. Commun. 59, 14745-14758 (2023).Shinoyama, M. et al. Cortical region-specific engraftment of embryonic stem cell-derived neural progenitor cells restores axonal sprouting to a subcortical target and achieves motor functional recovery in a mouse model of neonatal hypoxic-ischemic brain injury. Front. Cell. Neurosci. 7, (2013). Kriks, S. et al. Dopamine neurons derived from human ES cells efficiently engraft in animal models of Parkinson’s disease. Nature 480, 547-551 (2011). Bershteyn, M. et al. Human pallial MGE-type GABAergic interneuron cell therapy for chronic focal epilepsy. Cell Stem Cell 30, 1331-1350.e11 (2023). Davis, M. F. et al. Inhibitory Neuron Transplantation into Adult Visual Cortex Creates a New Critical Period that Rescues Impaired Vision. Neuron 86, 1055-1066 (2015). Kirkeby, A. et al. Preclinical quality, safety, and efficacy of a human embryonic stem cell-derived product for the treatment of Parkinson’s disease, STEM-PD. Cell Stem Cell 30, 1299-1314. e9 (2023). Falkner, S. et al. Transplanted embryonic neurons integrate into adult neocortical circuits. Nature 539, 248-253 (2016). Kim, T. W. et al. TNF-NF-KB-p53 axis restricts in vivo survival of hPSC-derived dopamine neurons. Ce / / 187, 3671-3689.e23 (2024). Brundin, P. et al. Improving the Survival of Grafted Dopaminergic Neurons: A Review over Current Approaches. Cell Transplant. 9, 179-195 (2000). Cunningham, M. et al. hPSC-Derived Maturing GABAergic Interneurons Ameliorate Seizures and Abnormal Behavior in Epileptic Mice. Cell Stem Cell 15, 559-573 (2014). Park, T.-Y. et al. Co-transplantation of autologous Treg cells in a cell therapy for Parkinson’s disease. Nature 619, 606-615 (2023). Azemi, E., Gobbel, G. T. & Cui, X. T. Seeding neural progenitor cells on silicon-based neural probes: Laboratory investigation. J. Neurosurg. 113, 673-681 (2010). Purcell, E. K., Seymour, J. P., Yandamuri, S. & Kipke, D. R. In vivo evaluation of a neuralstem cell-seeded prosthesis. J. Neural Eng. 6, 026005 (2009). Tang-Schomer, M. D. et al. Film-Based Implants for Supporting Neuron-Electrode Integrated Interfaces for The Brain. Adv. Fund. Mater. 24, 1938-1948 (2014). Serruya, M. D. et al. Engineered Axonal Tracts as “Living Electrodes” for Synaptic- Based Modulation of Neural Circuitry. Adv. Fund. Mater. 28, 1701183 (2018). Dayo O. Adewole1,2,3,4, Laura A. Struzyna1 ,2,3,4, Justin C. Burrell 1,2, 3, James P.Harrisl ,2, Ashley D. Nemes1 ,2, Dmitriy Petrov1,2, Reuben H. Kraft5, H. lsaac Chen1 ,2, Mijail D. Serruya2,6,7,8, John A. Wolf1,2, D. Kacy Cullenl ,2,3,4*. Development of optically controlled “living electrodes” with long-projecting axon tracts for a synaptic brain-machine interface. Sci. Adv. (2021). Sifringer, L. et al. An implantable biohybrid nerve model towards synaptic deep brain stimulation. Preprint at https: / / doi.org / 10.1101 / 2024.05.31.596665 (2024). Dong, X. et al. Human cerebral organoids establish subcortical projections in the mouse brain after transplantation. Mol. Psychiatry 26, 2964-2976 (2021). Mansour, A. A. et al. An in vivo model of functional and vascularized human brain organoids. Nat. Biotechnol. 36, 432-441 (2018). Revah, O. et al. Maturation and circuit integration of transplanted human cortical organoids. Nature 610, 319-326 (2022). Daviaud, N., Friedel, R. H. & Zou, H. Vascularization and Engraftment of Transplanted Human Cerebral Organoids in Mouse Cortex. eneuro S, ENEURO.0219-18.2018 (2018). Rochford, A. E. et al. Functional neurological restoration of amputated peripheral nerve using biohybrid regenerative bioelectronics. Sci. Adv. 9, eadd8162 (2023). Irwin, Z. T. et al. Chronic recording of hand prosthesis control signals via a regenerative peripheral nerve interface in a rhesus macaque. J. Neural Eng. 13, 046007 (2016). Urbanchek, M. G. et al. Development of a Regenerative Peripheral Nerve Interface for Control of a Neuroprosthetic Limb. BioMed Res. Int. 2016, 1-8 (2016).Bryson, J. B., Kourgiantaki, A., Jiang, D., Demosthenous, A. & Greensmith, L. An optogenetic cell therapy to restore control of target muscles in an aggressive mouse model of Amyotrophic Lateral Sclerosis. Preprint at https: / / doi.Org / 10.7554 / eLife.88250.2 (2023). Stieglitz, I, Ruf, H. H., Gross, M., Schuettler, M., & Meyer, J. U. A biohybrid system to interface peripheral nerves after traumatic lesions: design of a high channel sieve electrode. Biosens. Bioelectron. 178 685-696 Httpsdoiorg 101016s0956-56630200019-2. Lindstrom, S. & Andersson-Svahn, H. Single-Cell Culture in Microwells, in Single-Cell Analysis (eds. Lindstrom, S. & Andersson-Svahn, H.) vol. 85341-52 (Humana Press, Totowa, NJ, 2012). Swennenhuis, J. F. et al. Self-seeding microwell chip for the isolation and characterization of single cells. Lab. Chip 15, 3039-3046 (2015). Huber, D. et al. Sparse optical microstimulation in barrel cortex drives learned behaviour in freely moving mice. Nature 451 , 61-64 (2008). Luis-Islas, J., Luna, M., Floran, B. & Gutierrez, R. Optoception: Perception of Optogenetic Brain Perturbations. eneuro S, ENEURO.0216-22.2022 (2022). Prsa, M., Gali anes, G. L. & Huber, D. Rapid Integration of Artificial Sensory Feedback during Operant Conditioning of Motor Cortex Neurons. Neuron 93, 929-939.e6 (2017). Ragunathan Padmashri, Kevin Tyner, Anna Dunaevsky. Implantation of a Cranial Window for Repeated In Vivo Imaging in Awake Mice. J. Vis. Exp. Hilgenberg, L. G. W. & Smith, M. A. Preparation of Dissociated Mouse Cortical Neuron Cultures. J. Vis. Exp. 562 (2007) doi:10.3791 / 562. Spiegel, I. et al. Npas4 Regulates Excitatory-Inhibitory Balance within Neural Circuits through Cell-Type-Specific Gene Programs. Cell 157, 1216-1229 (2014). Rowitch, D. H. & Kriegstein, A. R. Developmental genetics of vertebrate glial-cell specification. Nature 468, 214-222 (2010). Hochbaum, D. R. et al. All-optical electrophysiology in mammalian neurons usingengineered microbial rhodopsins. Nat. Methods 11 , 825-833 (2014). Dana, H. et al. Sensitive red protein calcium indicators for imaging neural activity. eLife 5, e12727 (2016). Reimsnider, S., Manfredsson, F. P., Muzyczka, N. & Mandel, R. J. Time Course of Transgene Expression After Intrastriatal Pseudotyped rAAV2 / 1 , rAAV2 / 2, rAAV2 / 5, and rAAV2 / 8 Transduction in the Rat. Mol. Ther. 15, 1504-1511 (2007). Yuste, R. Neuronal ensembles: Building blocks of neural circuits. Neuron Volume 112, p875-892. Schuz, A. & Palm, G. Density of neurons and synapses in the cerebral cortex of the mouse. J. Comp. Neurol. 286, 442-455 (1989). Sortwell, C. E., Pitzer, M. R. & Collier, T. J. Time Course of Apoptotic Cell Death within Mesencephalic Cell Suspension Grafts: Implications for Improving Grafted Dopamine Neuron Survival. Exp. Neurol. 165, 268-277 (2000). Emgard, M., Karlsson, J., Hansson, O. & Brundin, P. Patterns of Cell Death and Dopaminergic Neuron Survival in Intrastriatal Nigral Grafts. Exp. Neurol. 160, 279-288 (1999). Thomas J. Mahalik, W.E. Hahn, Gerald H. Clayton, & Gregory P. Owens. Programmed Cell Death in Developing Grafts of Fetal Substantia Nigra. Exp. Neurol. Volume 129, Issue 1 , September 1994, Pages 27-36,. Abbott, N. J., Ronnback, L. & Hansson, E. Astrocyte-endothelial interactions at the blood-brain barrier. Nat. Rev. Neurosci. 7, 41-53 (2006). Knudsen, E. B. et al. A thin-film optogenetic visual prosthesis. Preprint at https: / / doi.org / 10.1101 / 2023.01.31.526482 (2023). Nemani, K. V, Moodie, K. L., Brennick, J. B., Su, A. & Gimi, B. In vitro and in vivo evaluation of SU-8 biocompatibility. Mater. Sci. Eng. C 33, 4453-4459 (2013). Stringer, C., Michaelos, M. & Pachitariu, M. Cellpose: a generalist algorithm for cellularsegmentation.58. Kronegg, J. Analysis of bit rate definitions for Brain-Computer Interfaces.

Claims

CLAIMSWe Claim:

1. A system configured to be implanted on a surface of a brain of a user, comprising:• a set of genetically modified cells;• a set of scaffolds, wherein the set of genetically modified cells are seeded between adjacent pairs of scaffolds in the set of scaffolds, wherein each scaffold in the set of scaffolds comprises:• an array of recording electrodes configured to receive signals from the set of genetically modified cells; and• an array of pLEDs configured to transmit light signals to the set of genetically modified cells based on control instructions; and• a controller communicatively connected to the set of scaffolds, wherein the controller is configured to:• receive data from the array of recording electrodes; and• determine control instructions for the array of pLEDs.

2. The system of Claim 1, wherein the genetically modified cells comprise genetically modified neurons derived from pluripotent stem cells.

3. The system of Claim 1, wherein the system further comprises a gel between adjacent pairs of scaffolds in the set of scaffolds, wherein the gel is configured to retain the genetically modified cells.

4. The system of Claim 3, wherein the gel is configured to support growth of axons of the genetically modified cells out of the gel.

5. The system of Claim 1, further comprising an anchor configured to secure the set of scaffolds to the surface of the brain of the user.

6. The system of Claim 1, wherein the controller comprises a proximal controller module and a distal controller module, wherein the proximal controller module is coupled to the set of scaffolds, and wherein the distal controller module is anchored to the skull of the user, wherein the distal controller module is communicatively coupled to an external device.

7. The system of Claim 6, further comprising a flexible connector connecting the proximal controller module to the distal controller module.

8. The system of Claim i, wherein, for each scaffold in the set of scaffolds, the array of recording electrodes and the array of pLEDs are located on opposing faces of the scaffold.

9. The system of Claim 1, wherein the genetically modified cells are transfected with a gene for a light-sensitive protein, wherein the genetically modified cells produce biochemical signals in response to receiving the light signals.

10. The system of Claim 1, wherein at least 100 genetically modified cells are seeded between each pair of adjacent scaffolds.

11. The system of Claim 1, wherein the set of scaffolds comprises at least 10 scaffolds.

12. A system configured to be implanted on a surface of a brain of a user, comprising:• a first scaffold comprising an array of pLEDs on a first face of the first scaffold;• a second scaffold comprising an array of recording electrodes on a first face of the second scaffold;• a spacer separating the second scaffold from the first scaffold, the spacer positioned between the first face of the first scaffold and the first face of the second scaffold;• a gel positioned between the array of pLEDs and the array of recording electrodes; and• a set of genetically modified cells retained within the gel, wherein the array of recording electrodes is configured to receive signals from the set of genetically modified cells, wherein the array of pLEDs is configured to transmit light signals to the set of genetically modified cells.

13. The system of Claim 12, wherein the spacer comprises a first raised conductive element in contact with the first face of the first scaffold and second raised conductive element in contact with the first face of the second scaffold, the system further comprising a connector connecting the first raised conductive element to the second raised conductive element.

14. The system of Claim 12, wherein the first scaffold further comprises a second array of recording electrodes on a second face of the first scaffold, wherein thesecond face of the first scaffold is opposite the first face of the first scaffold, and wherein the second scaffold further comprises a second array of pLEDs on a second face of the second scaffold, wherein the second face of the second scaffold is opposite the first face of the second scaffold.

15. The system of Claim 12, wherein the set of genetically modified cells comprise axons extending out of the gel, wherein the axons interface with native neurons in the brain of the user.

16. The system of Claim 12, wherein a thickness of the spacer is at least 100pm.

17. The system of Claim 12, wherein the genetically modified cells are transfected with a gene for a light-sensitive protein, wherein the genetically modified cells produce biochemical signals in response to receiving the light signals.

18. The system of Claim 17, wherein the set of cells comprise hypoimmune cells, wherein the set of cells are further transfected with a killswitch gene.

19. The system of Claim 12, wherein the array of pLEDs comprises at least 100 pLEDs.

20. The system of Claim 12, wherein the array of recording electrodes comprises at least 100 recording electrodes.

Images