Low-cost Neuronal Network Electronic Readout Circuit for Micro-Graphitic Diamond Multi-Electrode Arrays

The development of advanced microelectrode array (MEA) technology interfaced with low-noise high-gain conditioning circuits is crucial for evaluating the electrophysiological properties of neurons. The simultaneous and parallel access to the activity of a large population of neurons within a network aids the investigation of information processing and learning aspects of neural behavior. This project presents an early-stage design phase of a low-cost electronic readout circuit interfaced with Micro-Graphitic Diamond Multi-Electrode Arrays (µG-D-MEAs), for the real-time monitoring of the activity of cultured neurons in a potentiometric configuration. The developed Analog Front-End (AFE) is a single-channel device consisting of multiple amplifying and filtering stages. The input signal’s amplitude is of the order of tens of µV, thus the AFE has been designed to have a high gain and match the very low noise requirements imposed by the neuronal cell cultures. The most critical amplifying stage is based on the LMP7721 integrated circuit (a precision operational amplifier) which exhibits an ultra-low input bias current of 3 fA and voltage noise of 6.5 nV/&#8730;Hz. The AFE has been assessed by simulating a neuron spike shape by using the Agilent B2912A Precision Source/Measure Unit. It exhibited good performance when compared with a real neuron spike recorded with the commercial MEA1060-Inv-BC from Multichannel Systems, resulting in a <10 µVRMS background noise. Finally, dopaminergic neurons from substantia nigra (SN) have been used both as a test bench for the AFE and for scientific research. The device has been coupled with the cell culture inside an incubator with a controlled temperature of 37°C and the signals have been acquired from all 60 channels. The recordings showed the presence of the neuron spikes with a 2.82 Hz mean firing frequency and a &#8722;31.48 µV average peak amplitude. The noise level varied depending on the selected channel and, in the worst case, it proved to be smaller than 10 µVRMS. The results have been compared with the commercial device’s acquisitions, showing similar performances. The outcomes demonstrated that this project is ideal to be the first step of a wider plan. Looking to the future, further developments of the device may lead to extending the number of recording channels and to integrating the amperometric configuration with the potentiometric one, for a dual simultaneous and parallel mode.