Advances in Electrical and Computer Engineering Volume 20, Number 3, 2020 A Digital Signal Amplification Device for Microelectrode Arrays based on Stochastic Resonance Francisco FAMBRINI 1 , João Batista DESTRO-FILHO 2 , Luis Mariano Del Val CURA 3 , Diego SAQUI 1,4 , José Hiroki SAITO 1,3 1 Computation Department, UFSCar – Federal University of São Carlos, São Carlos, Brazil 2 UFU – Federal University of Uberlândia, Uberlândia, Brazil 3 UNIFACCAMP – University Center of Campo Limpo Paulista, Campo Limpo Paulista, Brazil 4 IFSULDEMINAS – Federal Institute of Sul of Minas Gerais, Muzambinho, Brazil djhs@ufscar.br 1 Abstract—In this work, an experimental study was carried out about the construction of an amplification equipment based on the phenomenon of stochastic resonance (SR), which was initially thought to detect spikes and bursts from human and animal neuronal tissue, both in vitro (from microelectrode array, MEA) and in vivo, from electrodes in the cerebral cortex of mammals. The implemented equipment was called CADSR (Computer-Aided Digital Stochastic Resonator) and brings as innovation the fact of being controlled and monitored by the computer, through a graphical interface that allows an automatic tuning, making it possible to obtain the optimum level of noise to maintain SR in real-time. Experimental results show that for electrical signals from multi-electrode arrays with amplitude below 25 microvolts, the amplification system using stochastic resonance is better than conventional amplifier systems, which use operational amplifiers in linear configurations. Index Terms—multielectrode, signal, stochastic, resonance, amplifier. I. INTRODUCTION A planar Microelectrode Array (MEA) of in-vitro neuron culture is used to record extracellular electrophysiological signals, with the possibility of chemical or electrical stimulation before or during the signal recording. Basically, the device is composed of a culture chamber, similar to a Petri dish, where the biological cell culture is placed on a number of microelectrodes. MEA is used to record signals from neuronal tissue slices or dissociated neuron cultures, allowing the investigation of electrical spontaneous or stimulated activities of neurons, with or without the use of drugs or medicines. Fig. 1(A) shows a MEA60 device from Multichannel Systems [1] used in this work, and in Fig. 1(B), its microelectrodes in the center are highlighted. Figure 1. (A) MEA60 . (B) Amplified details of microelectrodes in the center of MEA60 multielectrode array In this work experiments about the construction of MEA signal amplification equipment are realized based on the phenomenon of stochastic resonance (SR). Noise is generally detrimental to electronic systems because it limits performance, and great efforts are being made by physicists and engineers to reduce noise, such as filtering, feedback compensation, electromagnetic shielding, phase input, etc. However, in the last 40 years, several studies have demonstrated that many physical and biological systems work better in the presence of noise, such as biological neurons and electronic systems. Stochastic resonance is a term used to describe any phenomenon in which the presence of noise or noise applied on a nonlinear system improves the response to a particular input signal. This is a subject little known even by researchers in the area of Signal Processing. The SR occurs only in systems whose output presents non-linearity in relation to the applied input signal. An example of SR application is on static, dynamic, and functional balance in the elderly and in patients with neurodegenerative diseases [2]. A review of the applications of SR in rotating machine fault detection was described by Lu, He, and Wang [3]. A review of the impact of galvanic vestibular stimulation-induced SR on the output of the vestibular system was provided by Stefani, and Serrador [4]. Another interesting example of the application of the SR phenomenon in an electronic device was presented in the excellent article by Chiga, Tanaka, Yamazato, Tadokoro and Arai [5]. In this paper, the authors constructed a receiver for very weak signals of RF (Radio Frequency) based on Schmitt trigger electronic circuit, exemplifying the use of SR in wireless receiver systems [5]. Jayram and Mekuria described SR interference managing ontological cognitive radio for TV white space [6]. Another interesting paper is about a variant of SR termed Ghost Stochastic Resonance [7]. Some of the authors of this paper described in 2015 a very simple prototype of MEA signal amplification using SR, with only one channel, of an electronic circuit, generating white Gaussian noise, with amplitude adjusted manually [8], and in 2017, construction of a conventional system [9] of amplification of microelectrode signals, based on linear amplifiers, without the use of SR concept. In this paper it is presented an evolution of the first prototype described in [8], with the automatic detection of the 31 1582-7445 © 2020 AECE Digital Object Identifier 10.4316/AECE.2020.03004