Dynamics of a mutual inhibition between pyramidal neurons compared to human perceptual competition

Abstract

Visual perception emerges as the result of neural systems actively organizing intrinsically noisy visual signals. It is commonly assumed that selection processes of competing neurons underlie this emergence of perceptual organization. While the neural competition, realized by such a “mutual inhibition” circuit has been examined in many theoretical studies, its dynamic properties have not been investigated in real neurons. We have developed a “hybrid” system where two real-life pyramidal neurons in a mouse brain slice interact through a computer simulated mutual inhibition circuit. We found that simultaneous activation of the mutually inhibiting pyramidal neurons leads to bi-stable activity. We investigated the effects of noise and the effect of changes in the activation strength on the dynamics. We observed that the circuit exhibits dynamics strikingly similar to the known properties of bi-stable visual perception.

Augmenting Flexibility: Mutual Inhibition Between Inhibitory Neurons Expands Functional Diversity

“…On the other hand, many studies have examined inhibition in a more general way. Recurrent inhibition (either feedback or mutual inhibition) increases the number of basins of attraction [6], allowing the network to perform functions such as winner-take-all decision [6,41,42,43], bistable perception [32,33], oscillations [44,13,12], associative memory [45,20] and grid formation [46]. Interestingly, we show a wide variety of these computations can be performed by introducing mutual inhibition into the network.…”

“…1a for how it is defined in this study). However, there are several pieces of evidence suggesting that mutual inhibition plays a more crucial role than previously thought [16, 28, 29, 30].…”

“…For example, feedback inhibition is composed of a interconnected pair of excitatory and inhibitory neuron, which balances the network [26, 27, 36], allowing it to approximate arbitrary functions [17] and are useful in gain control. On the other hand, mutual inhibition increases the number of basin of attractions [5], allowing the network to perform functions such as winner-take-all decision [5, 37, 38, 39], bistable perception [28, 29], oscillations [40, 11, 10], associative memory [41, 16] and grid formation [42]. The wide variety of computations that mutual inhibition is capable of performing, along with connectome studies, all point to the idea that mutual inhibition is crucial in expanding a network’s functional repertoire.…”

“…These mutual inhibitory connections are found across species, from mammals (27,35,37) to simple organisms like flies (38). However, recent advances in multicellular recordings (35,(39)(40)(41) and simulation studies (19,39), has given rise to new evidence that suggests that faster mutual inhibition between inhibitory inter-neuron plays a more crucial role than previously thought.…”

“…For example, feedback inhibition is composed of an interconnected pair of excitatory and inhibitory neurons, which balances the network (30)(31)(32)46), allowing it to approximate an arbitrary motor sequence (14,33) and are useful in gain control. On the other hand, recurrent inhibition (either feedback or mutual inhibition) increases the number of basins of attraction (6), allowing the network to perform functions such as winner-take-all decision (6,(47)(48)(49)(50), bistable perception (39,40), oscillations (21,22,51), associative memory (19,52) and grid formation (53). This result, along with supporting connectome studies, all point to the idea that mutual inhibition is crucial in expanding a network's functional repertoire, by increasing and manipulating the number of basins of attraction.…”