Key result
CaCLEAN detection accuracy was more sensitive to RyR cluster density than the presence of mitochondria, with recall and precision varying between 0.69 and 0.82 depending on cluster packing.
Why the study?
Whether RyR clusters identified by CaCLEAN analysis of cardiomyocyte calcium fluorescence imaging were true or false positives remained unvalidated due to the absence of ground truth values.
The CaCLEAN algorithm's accuracy in detecting RyR clusters is highly sensitive to cluster density and distance from the imaging plane, with precision and recall ranging from 0.69 to 0.82.
Answers are grounded in this paper and cite their sources.
Accuracy varies with RyR cluster density; leaves open CaCLEAN applicability in intact cardiomyocytes pending validation.
In cardiomyocytes, the coordinated release of calcium ions from intracellular stores through ryanodine receptor (RyR) clusters is key to the generation of a calcium transient and induction of contraction. Recently, a deconvolution algorithm from radio astronomy was adapted and applied to analysis of calcium fluorescence imaging of cardiomyocytes. The algorithm, CaCLEAN, showed potential in revealing both the spatial locations of RyR clusters and their functional response in living cells. However, whether the RyR clusters identified by CaCLEAN analysis of the imaging data were true or false positives remained unvalidated in the absence of ground truth values. In this work, a structurally realistic finite element model was developed to simulate reaction-diffusion of calcium emanating from RyR clusters during the rising phase (first 30 ms) of the calcium transient. The effect of two sets of factors were examined with the model: (1) the number and spacing of simulated RyR clusters and (2) the effect of mitochondria acting as barriers to diffusion. Confocal fluorescence microscopy images were simulated from the model results and analysed using CaCLEAN. The performance of CaCLEAN was found to be sensitive to cluster spacing and distance from the imaging plane. In a case with sparsely-packed clusters, detection recall and precision were 0.82 for clusters up to 610 nm from the imaging plane; in a densely-packed cluster case, recall and precision were 0.69 for clusters up to 280 nm from the imaging plane. Users interested in applying CaCLEAN to their data should therefore consider the likely density of cluster distributions and the trade-off between precision and recall when determining the maximum relevant depth of RyR clusters in their application.
No takes yet. Share an insight, caveat, or question.
Ladd et al. (2019) studied this question. CaCLEAN algorithm vs. Ground truth from computational model was evaluated on Detection recall and precision. CaCLEAN detection accuracy was more sensitive to RyR cluster density than the presence of mitochondria, with recall and precision varying between 0.69 and 0.82 depending on cluster packing.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: