| Literature DB >> 25128722 |
Dimitri Boiroux1, Yoshihiko Oke2, Fumikazu Miwakeichi3, Yoshitaka Oku4.
Abstract
In point scanning imaging, data are acquired by sequentially scanning each pixel of a predetermined area. This way of scanning leads to time delays between pixels, especially for lower scanning speed or large scanned areas. Therefore, experiments are often performed at lower framerates in order to ensure a sufficient signal-to-noise ratio, even though framerates above 30 frames per second are technically feasible. For these framerates, we suggest that it becomes crucial to correct the time delay between image pixels prior to analyses. In this paper, we apply temporal interpolation (or pixel timing correction) for calcium imaging in two-photon microscopy as an example of fluorescence imaging. We present and compare three interpolation methods (linear, Lanczos and cubic B-spline). We test these methods on a simulated network of coupled bursting neurons at different framerates. In this network, we introduce a time delay to simulate a scanning by point scanning microscopy. We also assess these methods on actual microscopic calcium imaging movies recorded at usual framerates. Our numerical results suggest that point scanning microscopy imaging introduces statistically significant time delays between image pixels at low frequency. However, we demonstrate that pixel timing correction compensates for these time delays, regardless of the used interpolation method.Entities:
Keywords: Calcium imaging; Computational neuroscience; Point scanning; Time delay correction
Mesh:
Substances:
Year: 2014 PMID: 25128722 DOI: 10.1016/j.jneumeth.2014.08.008
Source DB: PubMed Journal: J Neurosci Methods ISSN: 0165-0270 Impact factor: 2.390