Researchers design new open source technology to interface with living neurons

Neurons communicate in complex ways and respond to stimuli within a vast network, orchestrating essential functions from basic bodily processes to complex thoughts. Traditional neuroscience methods, relying on in vivo electrophysiology (within a living organism), often struggle to address the complexity of the brain as a whole.

An alternative approach is to extract cells from the body and conduct studies in a culture dish (in vitro), providing researchers with greater control and precision in measuring neuronal processes.

In a new study presented in Advanced scienceresearchers unveil a cost-effective, open-source in vitro system for interfacing with neurons, providing a more accessible route for researchers interested in neuronal interactions.

The study is part of a larger project called Mind in Vitro, which explores how neurons interact with each other, not only to better understand the functions underlying complex systems like the brain, but also with the aim of eventually using in vitro neural networks for calculation. . To achieve such a goal, the MiV project encompasses a group of highly interdisciplinary researchers, including those from computer science, engineering, neurobiology, physiology, etc.

“The goal of the MiV project is to eventually use neurons for computation,” explained Zhi (Andrew) Dou, a graduate student in the Gazzola lab who helped lead the project. “This would create a system that is dynamic and constantly evolving, unlike traditional IT systems, and which would also be more energy efficient.”

The new MiV study, led by Mattia Gazzola (M-CELS), associate professor of mechanical science and engineering, Xiaotian Zhang, postdoctoral researcher in Gazzola’s lab, and Dou, describes an innovative approach to measuring neuronal activity at using an array of microelectrodes. (MEA). Although commercial systems using similar technology exist, they are often expensive and marketed for specific experimental approaches.

“The problem with current neuron interfacing technology is that they are mostly commercial systems, which are more standardized for specific testing conditions, usually in traditional biochemistry or neuroscience,” Zhang said. “As we want to design computing paradigms from living cells, our motivation was to move away from more standardized commercial systems and build something ourselves over which we would have full control.”

In the new MiV device, cells are placed on a plate containing MEA, allowing the technology to interface with neuronal substrates. The electrodes detect the voltage coming from the neurons, and this voltage is amplified and sent to a computer, which then processes the data.

Compared to the standard commercial 60-electrode system, the MiV system has more than 500 electrodes, allowing more data to be collected at once. The system also incorporates improved and/or new features compared to commercial systems, such as portability, two-way communication with neurons, the ability to image neurons during recordings, and the ability to test multiple input types ( electrical and optical) and several cell populations at the same time.

Despite these improvements, the cost of building the MiV system is incredibly 10 times cheaper than the cost of a commercial system and fully customizable. To help spread the system, researchers have made their hardware and software models for the MiV system open source and freely available online.

“It was very important to us that our software and hardware be open source,” Zhang said. “We wanted to reduce costs to make this technology accessible to others, and make it more flexible and versatile so that researchers can truly customize their experimental settings based on their needs. Our system is not intended for just one laboratory or to a single group. people interested in a particular thing, it’s for everyone.”

“We designed our system to be very easy to manufacture and composed of inexpensive parts, which will allow many laboratories that cannot afford the commercial system to have their own system,” Dou said. “Even though the original design is intended for computational studies, we’ve made the structure easy to redesign and expand, so we’re pretty sure it will satisfy researchers, no matter what type of studies they want to do.”

The researchers say other labs are already showing interest in the new technology for their own experiments. The MiV team plans to continue improving the design to make the hardware more automated and accessible, as well as exploring ways to quantify and extend the technology’s performance.

Links to hardware designs and open source software can be found at and respectively.