From Single Neurons to Brain Waves: How Neural Population Dynamics Are Redefining Brain-Computer Interfaces 🧠⚡

In his recent presentation at BCI Samara 2026, leading neuroscientist Mikhail Lebedev unpacked the evolution of neurophysiology—and why the future of Brain-Computer Interfaces (BCIs) relies on shifting our focus from single-neuron spikes to complex neural population dynamics and traveling waves.

Here are the key takeaways from the talk:

1️⃣ The Era of Single-Neuron Tracking

Neurophysiology spent decades obsessed with single units—from Hubel & Wiesel’s visual orientation columns and Rizzolatti’s mirror neurons to category-selective neurons in the prefrontal cortex. While foundational, single-neuron firing rates only paint a fraction of the full picture.

2️⃣ The Scalability Bottleneck in BCIs

From early multi-electrode arrays (Nicolelis & Lebedev recording up to 2,000+ neurons simultaneously in primates) to modern high-density implants like Neuralink, channel counts have exploded. However, recording thousands of neurons introduces a massive dimensionality problem: how do we meaningfully decode all this data?

3️⃣ Rotational Dynamics vs. Traveling Waves

Popular modern frameworks (such as Shenoy & Churchland's rotational dynamics) reduce high-dimensional neural activity into low-dimensional trajectories. Lebedev’s research demonstrates that these "rotations" can often be explained by traveling waves—sequential, phase-delayed computational pulses sweeping across neural populations.

4️⃣ The Wave Revolution & Analog Computing

Top-down brain control operates at lower frequency oscillations, while bottom-up sensory inputs travel at higher frequencies. Incorporating spatial wave patterns and brain oscillations—rather than simple mean firing rates—will be critical for the next generation of AI-driven, invasive BCIs.

💡 The Big Takeaway:

Building higher-bandwidth BCIs isn't just about adding more electrodes; it requires decoding the brain's wave-like computational geometry.

https://youtu.be/YPd2r6sdSts?si=lejXcyUg3dAPkBLg

#Neuroscience #BCI #BrainComputerInterface #Neuralink #Neurotech #AI #BiomedicalEngineering #DeepTech