Understanding Networks of Computing Chemical Droplet Neurons Based on Information Flow

Abstract

In this paper, we present general methods that can be used to explore the information processing potential of a medium composed of oscillating (self-exciting) droplets. Networks of Belousov-Zhabotinsky (BZ) droplets seem especially interesting as chemical reaction-diffusion computers because their time evolution is qualitatively similar to neural network activity. Moreover, such networks can be self-generated in microfluidic reactors. However, it is hard to track and to understand the function performed by a medium composed of droplets due to its complex dynamics. Corresponding to recurrent neural networks, the flow of excitations in a network of droplets is not limited to a single direction and spreads throughout the whole medium. In this work, we analyze the operation performed by droplet systems by monitoring the information flow. This is achieved by measuring mutual information and time delayed mutual information of the discretized time evolution of individual droplets. To link the model with reality, we use experimental results to estimate the parameters of droplet interactions. We exemplarily investigate an evolutionary generated droplet structure that operates as a NOR gate. The presented methods can be applied to networks composed of at least hundreds of droplets.

Cancer classification with a network of chemical oscillators

“…The rapid production of activator yields oxidation of the catalyst, seen as the change of droplet color from red to blue and named below as an excitation phase. Bearing in mind the experimental results on interactions between droplets 43,47 we assumed that the durations of the excitation, refractory and responsive phases are 1 s, 10 s, and 19 s respectively. These numbers sum up to the typical oscillation period of 30 s. 43 When illumination is applied the droplet is switched into the refractory phase and remains in it.…”

On Boolean gates in fungal colony

“…They are based on interactions of fluid streams, signals propagating along conductors or excitation wave-fronts, see e.g. [67,64,10,10,81,76,37,29,30,70,27,73,70,31]. Typically, logical gates and their cascade implemented in an excitable medium are 'handcrafted' to address exact timing and type of interactions between colliding wavefronts [67,64,1,9,74,8,20,71,82,72,32,7,69].…”

Belousov–Zhabotinsky reaction in liquid marbles

“…These reactions have proved to be potential media for developing future and emergent computing devices based on the interaction of chemical wave-fragments. A substantial number of theoretical studies and experimental prototypes of computing devices have been implemented using this media; image processors and memory devices [15][16][17], logical gates implemented in geometrically constrained BZ media [18,19], approximation of the shortest path of excitation waves [20][21][22], information coding using the frequency of oscillations [23], onboard controllers for robots [24][25][26], chemical diodes [27], neuromorphic architectures [28][29][30][31][32][33][34] and associated memory [35,36], wave-based counters [37] and other information processors [32,[38][39][40].…”

“…To provide insights into these processes and the use of BZ media for the applications above, previous studies utilised the BZ reaction in spatially confining media such as thin hydrogel films [41], small droplets in microfluidic devices [32,[42][43][44][45][46][47][48][49][50][51][52], vesicles [53][54][55][56][57], emulsions [58], BZ-AOT emulsions [9-11, 17, 59-64], cation exchange resins [65][66][67][68][69][70][71] and 3D printed structures [42,72,73].…”