Fast PCA for processing calcium-imaging data from the brain of drosophila melanogaster

The calcium-imaging technique allows us to record movies of brain activity in the antennal lobe of the fruitfly Drosophila melanogaster, a brain compartment where information about odors is processed. For signal processing that scales up with the growing data sizes in imaging, we have developed an approximate Principal Component Analysis (PCA) for fast dimensionality reduction. The approach relies on selecting a set of relevant pixels from the movies based on a priori knowledge about the nature of the data, ensuring a high-quality approximation. Once in PCA space, we can efficiently perform source separation, e.g to detect biological signals in the movies and to remove artifacts.