LAUSR

The rapid progress of calcium imaging techniques has reached a point where the activity of thousands to tens of thousands of cells can be recorded simultaneously with single-cell resolution in a field-of-view (FOV) of about ten mm2. Consequently, there is a pressing need for developing automatic cell detection methods for large-scale image data. Several research groups have proposed automatic cell detection algorithms. Almost all algorithms can solve large-scale optimization problems for data, including hundreds of cells recorded from a conventional FOV at a resolution of 512 × 512 pixels, but the solution becomes more difficult as the data size increases beyond that. To handle large-scale data acquired with the latest large FOV microscopes, we propose a method called low computational cost cell detection (LCCD) that is based on filtering and thresholding. We compared LCCD with two other methods, constrained non-negative matrix factorization (CNMF) and Suite2P. We found that LCCD makes it possible to detect cells in artificial and actual data showing a high number density of cells within a shorter time and with an accuracy comparable to or better than those of CNMF and Suite2P. Moreover, LCCD succeeded in detecting more than 20,000 active cells from data acquired with the latest microscopy, called FASHIO-2PM, with a FOV of 3.0 mm × 3.0 mm.