Signal analysis / PLOS ONE, 2025
Finding events in neural recordings.
Time-series segmentation methods to recognize epileptiform patterns in microelectrode-array recordings.
The question
How can relevant events be recognized in neural recordings with computationally inexpensive methods?
The study
The paper presents ZdensityRODE and AMPDE, two approaches based on time-series segmentation. They were evaluated using epileptiform activity recorded from mouse hippocampus-cortex slices with microelectrode arrays.
The work focuses on detecting and classifying recorded events. It motivates further research toward closed-loop stimulation; seizure prediction remains an application to investigate.
Read the paper
Published article and full text in PLOS ONE · DOI
Data and experimental results on Zenodo
The traces above are synthetic conceptual illustrations. They are not recordings or examples of either algorithm’s output.