NIXSolutions: Mind in Vitro – A Cost-Effective Neural Network Interface

American researchers have introduced a groundbreaking platform for studying living neurons and their networks, aiming to make neural experimentation more accessible and cost-effective. Named Mind in Vitro (MiV), this innovation is detailed in a report published in Advanced Science.

Creating a Universal Interface

Led by Mattia Gazzola, a team from the University of Illinois and Indiana University developed MiV with the goal of establishing a universal and accessible platform for in vitro neural network experiments. The device’s body, meticulously laser-cut from acrylic plates, incorporates a flexible mechanism with vibration-isolating pads. This design supports biocompatible microelectrode arrays (MEAs), which are pivotal for recording electrical signals from neurons.

NIX Solutions

Versatile Experimental Capabilities

MiV stands out with its ability to accommodate MEA chips of various sizes. The bottom plate features sliding guides and 3D-printed stops to center and secure MEA chips, manufactured through photolithography in a standard clean room. The top plate contains a printed circuit board with a matrix of soldered spring contacts, allowing simultaneous recording from up to 512 channels.

The platform supports both electrical and optical stimulation, enhancing its versatility for different experimental requirements. Electrical stimulation utilizes the Stimjim stimulator via spring contacts, while optical stimulation employs 465-nanometer LEDs and Doric lenses lasers connected to the MEA chip through optical fibers. The system’s adaptability extends to short-term and long-term recording, with the device easily fitting into incubators of various sizes and functionalities.

Data Processing and Cloud Computing

A single 24-hour experiment with a 512-channel, 30-kilohertz recording generates approximately four terabytes of data. To efficiently handle this vast amount of information, the researchers developed a cloud computing system. The core of the system is a Python-based pipeline computing platform that offers maximum user flexibility. The software supports high-performance computing systems and integrates with various data processing packages, such as H5py, Aim UI, Jupyter server, PyInform/IDTxl, Globus APIs, NeuralEnsemble, Kilosort, and scikit-learn.

Validation through Testing

The MiV platform underwent rigorous testing on 2D mouse neuron cultures, rat brain slices, and 3D bioengineered neural tissue organoids. Additionally, researchers modified the platform to be compatible with a fluorescence microscope, allowing simultaneous recording of neuronal activity using calcium ion current indicators alongside electrophysiological observations.

To demonstrate the platform’s portability, ruggedness, and reliability, the researchers disassembled it and transported it nearly 300 kilometers from Urbana-Champaign to Bloomington. Assembled in just half an hour, the unit reproduced previous tests with remarkable accuracy, showcasing its practicality and robustness.

Cost-Effective Solution

During a year of testing, the researchers amassed about a thousand hours of recordings. NIXSolutions notes that the MiV platform offers an economical solution, with a 59-channel version costing approximately $2,500 and a 512-channel version priced at $12,000. According to the researchers, this is 10 to 25 times cheaper than commercially available systems. The development aligns with the project’s overarching goal, named Mind in Vitro, emphasizing its future application for calculations using living neurons.