Real-time detection of sharp-wave ripples with a closed-loop stimulation framework

Sharp-wave ripples (SWRs) are thought to play a crucial role in memory consolidation, learning, and the formation of long-term memories by facilitating the transfer of information from the hippocampus to neocortical networks. To understand the mechanisms governing SWRs, we have developed a closed-loop stimulation framework tailored for the real-time detection and suppression of SWRs, leveraging the ONIX electrophysiology acquisition system and BONSAI-RX acquisition software. As part of our Bonsai package, we developed a convolutional neural network (CNN) for detection of SWRs recorded with laminar probes and Neuropixels probes. We quantify the performance our detection algorithm on previously published datasets and compare with the state-of-the-art algorithms. Crucially, the implementation of our detection algorithm and closed-loop feedback system enables dynamic suppression of SWRs as they occur, thereby advancing our understanding of their role through real-time electrical or optogenetic neurostimulation. This novel framework not only investigates the underlying mechanisms of SWRs but also provides a means for immediate intervention, offering insights beyond conventional approaches and contribute to a better understanding of the neural mechanisms governing memory processes.