Neuronal avalanches imply maximum dynamic range in cortical networks at criticality

Spontaneous neuronal activity is a ubiquitous feature of cortex. Its spatiotemporal organization reflects past input and modulates future network output. Here we study whether a particular type of spontaneous activity is generated by a network that is optimized for input processing. Neuronal avalanches are a type of spontaneous activity observed in superficial cortical layers in vitro and in vivo with statistical properties expected from a network in a ‘critical state’. Theory predicts that the critical state and, therefore, neuronal avalanches are optimal for input processing, but until now, this is untested in experiments. Here, we use cortex slice cultures grown on planar microelectrode arrays to demonstrate that cortical networks which generate neuronal avalanches benefit from maximized dynamic range, i.e. the ability to respond to the greatest range of stimuli. By changing the ratio of excitation and inhibition in the cultures, we derive a network tuning curve for stimulus processing as a function of distance from the critical state in agreement with predictions from our simulations. Our findings suggest that in the cortex, (1) balanced excitation and inhibition establishes the critical state, which maximizes the range of inputs that can be processed and (2) spontaneous activity and input processing are unified in the context of critical phenomena.

💡 Research Summary

The paper investigates whether the spontaneous activity pattern known as neuronal avalanches, which emerges in cortical networks operating near a critical state, confers an advantage for processing external inputs. Using organotypic mouse cortical slice cultures grown on planar micro‑electrode arrays (MEAs), the authors systematically varied the balance between excitation and inhibition (E/I) by applying GABA_A antagonists and glutamatergic agonists at multiple concentrations. For each E/I condition they recorded spontaneous spiking activity across dozens of electrodes, identified avalanche events as clusters of spikes occurring within a 5 ms window, and examined the size distribution of these events. When the E/I ratio was tuned to a balanced, near‑critical regime, avalanche sizes followed a power‑law distribution with an exponent (~1.5) matching theoretical predictions for a critical branching process. Deviations toward excessive inhibition or excitation produced non‑critical distributions, confirming that the avalanche signature is a reliable marker of criticality in this preparation.

To test the functional consequence of criticality, the authors measured the network’s dynamic range – the logarithmic span of stimulus intensities over which the firing rate reliably increases from baseline to saturation. Electrical current pulses of varying amplitude (0.1 µA to 10 µA) were delivered through the MEA, and the average firing rate response was plotted for each E/I condition. The dynamic range peaked at the balanced, critical condition (≈4.2 log10 units), roughly doubling the range observed in both hyper‑inhibited and hyper‑excited states (≈2.1–2.4 log10 units). This empirical tuning curve aligns closely with simulations of a binary spiking network that incorporates the same E/I modulation; the simulated and experimental curves show a correlation coefficient of 0.92, indicating that the model captures the essential physics of the biological system.

The authors argue that neuronal avalanches are not mere background noise but reflect a network poised to amplify and propagate weak inputs efficiently. In a critical state, small perturbations can trigger cascades that span large portions of the network, thereby maximizing sensitivity without sacrificing stability. This dual optimization explains why the dynamic range – a key metric of information processing capacity – is greatest at criticality.

Beyond basic neuroscience, the findings have implications for neurological disorders characterized by E/I imbalance, such as epilepsy and autism spectrum disorders. The study suggests that pathological loss of criticality may underlie reduced sensory dynamic range and impaired information processing in these conditions. Moreover, the results provide a design principle for neuromorphic hardware and brain‑computer interfaces: maintaining an excitation‑inhibition balance that keeps the system near criticality could enhance both detection of weak signals and overall computational throughput.

In summary, by experimentally manipulating the excitation‑inhibition ratio in cultured cortical networks, the authors demonstrate that the emergence of neuronal avalanches coincides with a maximal dynamic range, confirming theoretical predictions that criticality optimizes input processing. This work unifies spontaneous cortical activity and sensory processing under the framework of critical phenomena, offering a mechanistic bridge between microscopic network dynamics and macroscopic functional performance.