Semi-automated program for analysis of local Ca spark release with application for classification of heart cell type.

Local Ca spark releases are essential to the Ca cycling process. Thus, they play an important role in ventricular and atrial cell contraction, as well as in sinoatrial cell automaticity. Characterizing their properties in healthy cells from different regions in the heart can reveal the basic biophysical differences among these regions. We designed a semi-automatic Matlab Graphical User Interface (called Sparkalyzer) to characterize parameters of Ca spark release from any major cardiac tissue, as recorded in line-scan mode with a confocal laser-scanning microscope. We validated the algorithm on experimental images from rabbit sinoatrial, atrial, and ventricular cells loaded with Fluo-4 AM. The program characterizes general image parameters of Ca transients and sparks: spark duration, which indicates for how long the spark provides Ca to the closed intracellular mechanisms (typical value: 25±1, 23±1, 26±1ms for sinoatrial, atrial, and ventricular cells, respectively); spark amplitude, which indicates the amount of Ca released by a single spark (1.6±0.1, 1.6±0.2, 1.4±0.1F/F for sinoatrial, atrial, and ventricular cells, respectively); spark length, which is the length of the Ca wavelets fired out of a row of ryanodine receptors (5±0.1, 5±0.2, 3.4±0.3μm for sinoatrial, atrial, or ventricular cells, respectively) and number of sparks (0.14±0.02, 0.025±0.01, 0.02±0.01 for 1μm in 1s for sinoatrial, atrial, and ventricular cells, respectively). This method is reliable for Ca spark analysis of sinoatrial, atrial, or ventricular cells. Moreover, by examining the average value of Ca spark characteristics and their scattering around the mean, atrial, ventricular and sinoatrial cells can be differentiated.