Advances in Microelectrode Array Data Analysis: Developing Methods for Characterizing Neuronal Network Activity in vitro

Electrophysiological research with microelectrode arrays (MEAs) has emerged as a pivotal approach for studying in vitro neuronal network activity by offering repeated non-invasive measurements of activity from a population of neurons at multiple sites simultaneously. Current developments in microfluidics technology, MEA manufacturing and the use of human pluripotent stem cells (hPSCs) have collaboratively enhanced in vitro neuronal models to better mimic the human brain. The acquired data contains important insights into the functional dynamics of neural circuits, with the potential to advance developmental studies, support pharmacological research and uncover mechanisms underlying neurological disorders. However, the obtained data has limited utility without proper processing using refined and accurate analysis tools.

This PhD thesis focuses on advancing the methodologies for analyzing MEA data, addressing key challenges related to detection and characterization of neuronal activity. It aims to develop novel methods, optimize existing tools, and establish a comprehensive analysis pipeline applicable to MEA data obtained with different recording configurations including standard MEA setups and custom 3- compartment microfluidic MEA plates. Simultaneously, the thesis addresses the matter of making the produced scientific data publicly available and demonstrates data curation activities. The data processed in the thesis also includes MEA recordings of more widely utilized rat cortical cultures compared to hPSC-derived cultures.

As the result, the thesis provides a detailed description of the MEA data analysis pipeline development. The emphasis was made on accurate detection of extracellular neuronal spikes, assessment of synchronous bursting and functional connectivity analysis. The assessment of multi-level synchronous activity patterns in 3- compartment MEA data was the key element in this work. In addition, a collection of MEA data and analysis codes was organized, reformatted, supplemented with the data descriptors and shared openly in accordance with the Findability, Accessibility, Interoperability, and Reusability (FAIR) principles to promote open science.

Finally, the thesis provides real-life applications of the achieved analysis tools in a context of functional comparison of rat cortical and hPSC-derived cultures and characterization of kainic acid-induced alterations. The developed tools have demonstrated the ability to yield novel findings and hold promise for uncovering new insights into the electrophysiological behavior of neuronal populations in vitro. The MEA data analysis pipeline is intended for use in developmental and pharmacological research, and for in vitro disease modeling studies.