'SpikeNburst and 'Nicespike: Advanced Tools for Enhancing and Accelerating In Vitro High-Density Electrophysiology Analysis

High-density multi-electrode arrays (HD-MEAs) enable the recording of in vitro neuronal activity with exceptional spatial and temporal resolution. However, analysing these extensive datasets presents challenges, such as artefact removal, spike sorting, and accurate assessments of neuronal synchronization. Here, we introduce a structured protocol for conducting comprehensive HD-MEA analyses using two Python-based tools: spikeNburst and nicespike. This protocol provides a scalable and systematic approach to processing HD-MEA recordings, ensuring efficient data handling and robust analytical outcomes. The spikeNburst tool incorporates advanced methodologies for spike train filtering, burst and network burst detection, and synchronization analysis. Complementing this, we have implemented a full analysis pipeline in the nicespike tool, featuring GPU-accelerated spike sorting via template matching with Kilosort, enabling accurate identification of neuronal units across multiple electrodes. This protocol ensures more precise analyses by reducing redundancy and overestimation inherent in single-channel approaches. Moreover, both tools offer graphical user and command-line interfaces, ensuring accessibility for diverse user needs. We validated our protocol on in vitro neuronal culture recordings, demonstrating its suitability to identify somatic and dendritic features of neuronal units, characterize bursting behaviour, and quantify synchronization at both unit and network levels. By addressing critical limitations of existing methods, spikeNburst and nicespike provide a robust, scalable, and user-friendly framework for HD-MEA data analysis, enhancing the study of neural network dynamics and single-cell activity in detail.