Thumbnails:
List:
Year:
Category:
Session:
Poster:
Getting poster data...
Alen Adiyev, Dmitry Molotkov, Hiroki Asari (European Molecular Biology Laboratory, Epigenetics and Neurobiology Unit, Rome, Italy)
Calcium imaging is a widely used technique to monitor neuronal population dynamics. While computational methods to process somatic image datasets have been well established, those for axonal images remain inefficient. In this study, we developed an analysis pipeline based on the Calcium Imaging Analysis (CaImAn) open-source software for segmenting the time-lapse image data recorded from retinal ganglion cell axons in the mouse superior colliculus. Specifically, we propose to employ a convolutional neural network (CNN) for quality control of the spatial footprints generated by CaImAn, which in most cases needs to be done manually by experts. The CNN model was trained on a dataset consisting of 7320 manually annotated spatial footprints from 11 different recordings. As a result, the model was able to effectively classify the footprints as “good” or “bad”, hence eliminating the time-consuming manual selection procedure.