Artificial neurons successfully communicate with living brain cells

Northwestern University engineers have achieved a significant breakthrough in bioelectronics, developing printed artificial neurons that can directly interact with and activate biological brain cells. This innovation moves beyond mere imitation, offering a new paradigm for interfacing electronic devices with living neural systems. The flexible, low-cost devices produce electrical signals that closely mirror those generated by natural neurons, enabling them to effectively stimulate brain tissue.

Groundbreaking Experiments Demonstrate Neural Compatibility

In rigorous laboratory experiments, slices of mouse brain tissue were exposed to these novel artificial neurons. The results were compelling: the synthetic neurons successfully triggered responses in the biological neurons, a critical step demonstrating a new level of compatibility and communication between artificial electronics and living neural networks. This achievement marks a substantial leap forward from previous artificial neuron technologies, which often produced signals too simplistic or incompatible to elicit meaningful biological reactions.

The research, published on April 15th in the prestigious journal Nature Nanotechnology, details how the team leveraged unique material properties and advanced printing techniques to create devices capable of mimicking the complex firing patterns of biological neurons. This intricate dance of electrical signals is fundamental to brain function, and replicating it opens up a vast landscape of potential applications.

A New Era for Brain Interfaces and Neuroprosthetics

The implications of this advancement are far-reaching, bringing researchers closer to creating electronics that can directly and seamlessly interface with the human nervous system. This opens doors to revolutionary applications in brain-machine interfaces (BMIs) and neuroprosthetics. Imagine implants that could restore lost sensory functions, such as hearing or vision, or enable individuals to regain control over paralyzed limbs. The ability of these artificial neurons to generate biologically relevant signals is key to achieving such sophisticated restorative technologies.

For individuals suffering from neurological conditions or injuries, this technology could offer unprecedented hope. The precise and responsive nature of these artificial neurons suggests a future where damaged neural pathways can be bypassed or augmented, restoring functionality and improving quality of life. The development is particularly timely, as the global population ages and the prevalence of neurodegenerative diseases and spinal cord injuries continues to rise, creating an increasing demand for advanced therapeutic solutions.

Towards Energy-Conscious Artificial Intelligence

Beyond medical applications, this breakthrough holds immense promise for the future of computing, particularly in the realm of artificial intelligence (AI). The current trajectory of AI development is heavily reliant on massive datasets and computationally intensive training processes, leading to a significant and growing problem of energy consumption.

Mark C. Hersam, the Walter P. Murphy Professor of Materials Science and Engineering at Northwestern’s McCormick School of Engineering and the lead author of the study, highlighted this critical challenge. "The world we live in today is dominated by artificial intelligence (AI)," Hersam stated. "The way you make AI smarter is by training it on more and more data. This data-intensive training leads to a massive power-consumption problem. Therefore, we have to come up with more efficient hardware to handle big data and AI."

Hersam elaborated on the brain’s unparalleled efficiency: "Because the brain is five orders of magnitude more energy efficient than a digital computer, it makes sense to look to the brain for inspiration for next-generation computing." The human brain, a marvel of biological engineering, performs complex computations using an estimated 20 watts of power, a stark contrast to the gigawatts consumed by large data centers powering modern AI. This fundamental difference in energy efficiency is what drives the pursuit of brain-inspired computing.

The Limitations of Silicon and the Brain’s Unique Architecture

Traditional silicon-based computing systems achieve their power by packing billions of identical transistors onto rigid, two-dimensional chips. While effective for many tasks, this approach has inherent limitations when it comes to replicating the nuanced and dynamic nature of biological intelligence. Each transistor on a silicon chip operates in a uniform manner, and once manufactured, the system’s architecture is largely fixed.

The brain, however, operates on fundamentally different principles. It is a heterogeneous system, comprised of diverse types of neurons, each with specialized functions. These neurons are interconnected in intricate, three-dimensional networks that are not static but are constantly adapting and reconfiguring. This plasticity, the ability to form and adjust connections as learning occurs, is crucial for cognitive functions.

"Silicon achieves complexity by having billions of identical devices," Hersam explained. "Everything is the same, rigid and fixed once it’s fabricated. The brain is the opposite. It’s heterogeneous, dynamic and three-dimensional. To move in that direction, we need new materials and new ways to build electronics."

Previous attempts to create artificial neurons have often fallen short. Many produced overly simplistic electrical signals, requiring vast networks of these devices to achieve any degree of complex behavior, which in turn escalated energy demands. The Northwestern team’s innovation lies in overcoming these limitations by embracing the brain’s inherent complexity and plasticity.

Printable Materials Unlock Brain-Like Signal Complexity

The key to the Northwestern team’s success lies in their utilization of soft, printable materials that more closely mimic the brain’s structural and functional characteristics. Their approach centers on the development of "electronic inks" composed of nanoscale flakes of molybdenum disulfide (MoS₂), a semiconductor material, and graphene, an excellent electrical conductor. These inks are deposited onto flexible polymer substrates using a high-precision aerosol jet printing technique.

A significant innovation in this work involved a novel approach to handling the polymer used in the inks. Historically, researchers viewed this polymer as a detrimental component that interfered with electrical performance and was typically removed after printing. In contrast, the Northwestern team ingeniously leveraged this polymer to enhance the device’s functionality.

"Instead of fully removing the polymer, we partially decompose it," Hersam detailed. "Then, when we pass current through the device, we drive further decomposition of the polymer. This decomposition occurs in a spatially inhomogeneous manner, leading to formation of a conductive filament, such that all the current is constricted into a narrow region in space."

This controlled decomposition process results in the formation of a narrow conductive path. When current flows through this path, it triggers a sudden, sharp electrical response that closely resembles the "firing" of a biological neuron. Crucially, the resulting artificial neuron is capable of generating a wide spectrum of electrical signals, including single spikes, continuous firing patterns, and more complex bursting behaviors. This multifaceted signal generation capability is a significant advancement, more accurately reflecting the diverse communication methods employed by real neurons.

The ability of each artificial neuron to produce more complex and varied signals means that fewer components are required to perform advanced computational tasks. This reduction in component count directly translates to a substantial improvement in computing efficiency, a critical factor for the future of AI.

Rigorous Testing on Live Neural Tissue

To validate the functional compatibility of their artificial neurons with living biological systems, the Northwestern researchers collaborated with Professor Indira M. Raman, a leading neurobiologist at Northwestern University Feinberg School of Medicine. Raman’s team applied the signals generated by the artificial neurons to slices of mouse cerebellum, a brain region critical for motor control and coordination.

The experimental results provided strong evidence of the artificial neurons’ efficacy. The electrical spikes produced by the synthetic devices were found to match key biological properties, including their precise timing and duration. These signals were shown to reliably activate real neurons and trigger neural circuits in a manner analogous to natural brain activity.

"Other labs have tried to make artificial neurons with organic materials, and they spiked too slowly," Hersam noted, emphasizing the temporal precision achieved. "Or they used metal oxides, which are too fast. We are within a temporal range that was not previously demonstrated for artificial neurons. You can see the living neurons respond to our artificial neuron. So, we’ve demonstrated signals that are not only the right timescale but also the right spike shape to interact directly with living neurons." This precise temporal and waveform matching is crucial for effective communication within neural networks.

Sustainable Manufacturing and Profound AI Implications

Beyond their remarkable performance, the newly developed artificial neurons offer significant advantages in terms of manufacturing sustainability and cost-effectiveness. The printing process is straightforward and inexpensive, and the additive nature of aerosol jet printing ensures that material is deposited only where it is needed, minimizing waste. This eco-friendly approach aligns with growing calls for more sustainable technological development.

The energy efficiency aspect of this research is particularly pertinent given the escalating power demands of artificial intelligence. As AI systems become more sophisticated and data-intensive, the energy footprint of data centers is becoming a major global concern. These facilities already consume vast amounts of electricity, contributing to carbon emissions and placing a strain on energy grids. Furthermore, the significant water required for cooling these data centers presents a growing challenge to water resources.

Hersam articulated the urgency of addressing this power consumption issue: "To meet the energy demands of AI, tech companies are building gigawatt data centers powered by dedicated nuclear power plants," he stated. "It is evident that this massive power consumption will limit further scaling of computing since it’s hard to imagine a next-generation data center requiring 100 nuclear power plants. The other issue is that when you’re dissipating gigawatts of power, there’s a lot of heat. Because data centers are cooled with water, AI is putting severe stress on the water supply. However you look at it, we need to come up with more energy-efficient hardware for AI."

The research, titled "Multi-order complexity spiking neurons enabled by printed MoS₂ memristive nanosheet networks," was made possible through support from the National Science Foundation, underscoring the importance of foundational research in driving technological innovation. This work represents a critical step towards realizing more intelligent, efficient, and sustainable computing systems, bridging the gap between biological and artificial intelligence. The ability to print complex neural circuits at low cost and with minimal waste suggests a future where advanced bioelectronic devices are not only powerful but also accessible and environmentally responsible. The study’s findings are poised to influence the trajectory of AI hardware development, potentially leading to a new generation of computing that operates with the efficiency and adaptability of the human brain.