Random-forest segmentation and spatial analysis of injected cardiac spheroids in optically cleared myocardium
Accurate quantification of transplanted cardiac spheroids requires three-dimensional localisation within intact myocardium, yet this remains technically challenging. Optical clearing and light-sheet microscopy enable volumetric imaging of injection sites, but automated segmentation is difficult when transplanted spheroids and host tissue are labelled with the same fluorescent markers and cannot be separated by simple thresholding. We developed a random forest based pixel classification workflow for 3D detection of injected hiPSC derived cardiomyocyte and H9c2 spheroids in optically cleared rabbit myocardium. A supervised classifier trained on intensity, edge, and texture features generated a segmentation then grouped pixels via connected component analysis to reconstruct individual spheroids. The method showed good agreement with manual annotation and enabled automated extraction of spheroid size and spatial metrics. This accessible workflow enables reproducible three-dimensional quantification of transplanted spheroids in large light-sheet microscopy datasets and provides a practical route from volumetric imaging to spatial metrics in cardiac regeneration studies.