Automated measurement of cardiomyocyte monolayer contraction using the Exeter Multiscope

We apply a novel microscope architecture, the Exeter Multiscope, to the problem of acquiring image data in rapid succession from nine wells of a 96 well plate. We demonstrate that the new microscope can detect contraction in cardiomyocyte monolayers which have been plated into these wells. Each well is sampled using 500 x 500 pixels across a 1.4 x 1.4 mm field of view, acquired in three colours at 3.7 Hz per well. The use of multiple illumination wavelengths provides post-hoc focus selection, further increasing the level of automation. The performance of the Exeter Multiscope is benchmarked against industry standard methods using a commercial microscope with a motorised stage and demonstrates that the Multiscope can acquire data almost 40 times faster. The data from both Multiscope and the commercial systems are processed by a 'pixel variance' algorithm that uses information from the pixel value variability over time to determine the timing and amplitude of tissue contraction. This algorithm is also benchmarked against an existing algorithm that employs an absolute difference measure of tissue contraction.