Superpositions between non linear intermittency maps, application in biological neurons networks

This paper demonstrates that superposing and coupling non-linear intermittency maps to model critical and tricritical points generates robust biological-type spike trains, suggesting a potential framework for investigating neurological disorders caused by spike loss.

The Gist

The Big Idea: Turning Math into Brain Sparks

Imagine your brain is a massive orchestra. Usually, when we think of neurons (brain cells), we think of them firing like individual drummers hitting a beat. But this paper suggests that the brain works more like a giant, chaotic choir where thousands of voices blend together to create a specific kind of "music" that allows us to think.

The author, Yiannis Contoyiannis, is asking a very specific question: If you mix together thousands of these chaotic "voices," does the brain still produce the clear, sharp "sparks" (spikes) it needs to function, or does the noise drown them out?

The Ingredients: Two Types of Chaos

To understand the experiment, we first need to understand the two "ingredients" the author is mixing. In the world of physics, there are two special states of chaos that act like the "on" and "off" switches for a neuron:

1. The "Critical" Map (The Rising Spark): Think of this as a ball rolling up a hill. It moves slowly at first (a quiet period), then suddenly gains speed and shoots up (a burst). In the brain, this represents the excitation phase, where a neuron gathers energy to fire.
2. The "Tricritical" Map (The Falling Spark): This is the opposite. It's like a ball rolling down a steep hill. It starts high and crashes down quickly. In the brain, this represents the inhibition phase, where the neuron resets and cools down after firing.

The Discovery: In previous work, the author found that if you take one "Rising" map and one "Falling" map and tie them together, they create a perfect Spike Train—a pattern that looks exactly like the electrical signals in a real biological brain.

The Experiment: The Superposition Soup

The big question in this paper is: What happens if we don't just use one pair, but mix 10, 50, or even 100 of these maps together?

Imagine you are making a smoothie.

• Scenario A: You blend one strawberry and one banana. It tastes great (a perfect spike).
• Scenario B: You throw in 100 strawberries and 100 bananas. Does it still taste like a strawberry-banana smoothie, or does it turn into a muddy, unrecognizable sludge?

The author ran a computer simulation (a "numerical experiment") to see what happens when you "superimpose" (stack) these maps on top of each other, just like a microphone picking up the sound of 100 neurons at once.

The Results:

• The Good News: Even with 100 maps mixed together, the system still produced the "Power Law" (a specific mathematical signature of chaos) that indicates the system is still alive and critical. The "soup" didn't turn into static; it kept its chaotic rhythm.
• The Bad News (The Catch): While the rhythm remained, the shape of the spikes changed.
  • With 10 maps, the spikes were sharp and distinct, like clear drumbeats.
  • With 100 maps, the spikes started to blur. They became shorter, flatter, and sometimes disappeared entirely, turning into flat lines or "orthogonal structures" (like a plateau instead of a peak).

The Metaphor: The Traffic Jam of Thoughts

Think of the brain's information flow as cars driving on a highway.

• A Spike is a car zooming past at high speed.
• When you have a few cars (10 neurons), they zoom by clearly. You can see each one.
• When you have too many cars (100 neurons) all trying to zoom at once, they start to bump into each other. They merge into a single, slow-moving traffic jam. The individual cars (spikes) disappear, and you just see a long, flat line of traffic.

Why This Matters for Neurological Diseases

This is the most important part of the paper. The author suggests a scary possibility for human health:

If the brain gets "too crowded" with neurons firing at once, the sharp spikes needed for thinking might get smoothed out.

• Sharp Spikes = Clear Thoughts: To process information quickly and think clearly, your brain needs distinct, sharp electrical spikes.
• Flat Lines = Confusion: If the number of neurons in a specific area increases too much (or they get too coupled), the spikes might flatten out. The brain loses its ability to send distinct signals.

The author proposes that this "flattening" of spikes could be the root cause of neurological decline and loss of thinking ability. It's not that the brain stops working; it's that the signals become so overlapped that they can't be distinguished anymore. It's like trying to have a conversation in a room where everyone is shouting at the same time; the message is lost in the noise.

The Conclusion: Electrical Synapses are the Mixer

Finally, the paper points to electrical synapses (tiny bridges between neurons) as the place where this mixing happens. These bridges allow neurons to "superimpose" their signals instantly.

In a nutshell:
The brain is a delicate balance. It needs the chaos of many neurons working together to create the "music" of thought. But if that chaos gets too crowded, the music turns into noise, and the sharp "spikes" of intelligence turn into flat lines. This mathematical model gives us a new way to look at why our thinking might slow down as we age or when we suffer from brain diseases.

Technical Summary

1. Problem Statement

The paper addresses the challenge of modeling the complex dynamics of biological neural networks using non-linear chaotic systems. Specifically, it investigates whether the biological-type spike train (ST) generation mechanism—previously demonstrated by coupling a single "critical" intermittency map with a single "tricritical" map—remains valid when the system is scaled up.

The core research questions are:

• Can the superposition of multiple critical or tricritical intermittency maps (representing a field of neurons) preserve the underlying power-law distribution of laminar lengths (waiting times) characteristic of critical and tricritical states?
• Does the coupling of these superposed fields (representing the interaction between excitatory and inhibitory dynamics) still produce biological-type spike trains as the number of maps (neurons) increases?
• How does increasing the number of coupled neurons affect the fidelity of the spike train, and can this model explain neurological decline due to a loss of spike activity?

2. Methodology

The study employs a numerical simulation approach based on Type I intermittency maps derived from the Method of Critical Fluctuations (MCF).

• Mathematical Models:

  • Critical Map (Eq. 1): Represents second-order phase transitions (excitatory dynamics).
    $$\Phi_{n+1} = \Phi_n + u_1 \Phi_n^z + \epsilon_n$$
    Where $u_1 > 0$, $z > 0$, and $\epsilon_n$ is uniform noise. This map drives values upward (rise).
  • Tricritical Map (Eq. 3): Represents the crossover to first-order transitions (inhibitory dynamics).
    $$\Phi_{n+1} = \Phi_n - u_2 \Phi_n^{-z} + \epsilon_n$$
    Where $u_2 > 0$. The negative sign and negative exponent drive values downward (fall).
  • Power-Law Analysis: The distribution of laminar lengths ($L$) is analyzed using the fitting function $f(L) \sim L^{-p_2} e^{-p_3 L}$.
    • Critical State: $p_2 \in [1, 2)$.
    • Tricritical State: $p_2 \in [0.66, 1)$.
    • $p_3 \approx 0$ indicates proximity to the critical/tricritical point.

• Superposition Strategy:
  The authors simulate the superposition of $k$ maps (where $k=10$ and $k=100$).

  • Critical Superposition (Eq. 8): Summing $k$ critical maps with varying parameters ($u_i, z_i$).
  • Tricritical Superposition (Eq. 9): Summing $k$ tricritical maps with varying parameters.
  • Parameters ($z$) were varied randomly within integer intervals (e.g., $z \in [2, 6]$) to simulate heterogeneous neuron populations.

• Coupling Mechanism:
  The superposed critical time series ($X$) and tricritical time series ($Y$) are coupled. To ensure the tricritical component effectively drives the system down (hyperpolarization), the authors invert the sign of the tricritical series before coupling:

  • Coupling Scheme: $X(i) \leftrightarrow -Y(i)$.
  • The coupling defines the "laminar region" (grass/relaxation) and the "bursts" (spikes).

3. Key Results

A. Preservation of Criticality/Tricriticality in Superposition

• 10 Maps: Superpositions of 10 critical maps yielded a laminar length distribution exponent $p_2 = 1.20$ (Critical). Superpositions of 10 tricritical maps yielded $p_2 = 0.975$ (Tricritical).
• 100 Maps: Even with 100 maps, the system preserved its state.
  • 100 Critical maps: $p_2 = 1.15$ (Critical).
  • 100 Tricritical maps: $p_2 = 0.77$ (Tricritical).
• Conclusion: The superposition of non-linear intermittency maps preserves the power-law distribution of laminar lengths, maintaining the system in a critical or tricritical state regardless of the number of components, provided parameters are varied.

B. Spike Train Generation and Degradation

• Coupling Output: Coupling the superposed fields produced time series resembling biological membrane potentials, featuring excitation thresholds, hyperpolarization, and relaxation periods ("grass").
• Effect of Scale (10 vs. 100 Neurons):
  • 10 Maps: In a specific segment, 8 out of 9 structures were distinct spikes (88% spike fidelity).
  • 100 Maps: In a comparable segment, only 17 out of 39 structures were distinct spikes (44% spike fidelity).
• Observation: As the number of coupled maps increases, spikes begin to overlap, causing the sharp "edges" of the spikes to flatten into planes. The distinct spike structure is partially lost due to the interference of multiple simultaneous signals.

4. Key Contributions

1. Generalization of Intermittency Models: The paper demonstrates that the critical and tricritical dynamics are robust properties that survive the superposition of multiple heterogeneous maps, validating the use of these models for large-scale neural fields rather than single neurons.
2. Mechanism for Biological Spike Trains: It confirms that the coupling of critical (Na-ion pump dynamics) and tricritical (K-ion pump dynamics) fields can generate biological-type spike trains even in complex, superposed systems.
3. Quantitative Link to Neurological Decline: The study provides a theoretical framework suggesting that neurological diseases and cognitive decline may result from the "smearing" of spike trains. As the number of neurons in a network increases without proper synchronization, the spike fidelity drops (overlap), reducing the information transfer capacity of the brain.
4. Biological Interpretation of Synapses: The authors propose that electrical synapses (gap junctions) are the biological substrate for the superposition and coupling mechanisms observed in the model, as they allow for bidirectional coupling and synchronous network activity.

5. Significance

This work bridges Statistical Physics and Neuroscience. By mapping the mathematical concepts of criticality and tricriticality to the dynamics of biological neurons, the authors offer a new perspective on how brain networks process information.

• Theoretical Impact: It suggests that the brain operates near critical points to maximize information processing, but this state is fragile.
• Clinical Relevance: The finding that increasing network density (without adjusting timing) leads to spike degradation offers a potential explanation for cognitive decline and neurological disorders where signal clarity is lost. It implies that therapeutic strategies might need to focus on optimizing the timing (intervals) of neural firing to maintain spike distinctness in large networks.
• Future Applications: The model opens avenues for simulating neurological pathologies and designing artificial neural networks that mimic the chaotic yet structured dynamics of biological systems.