The genetic and phenotypic heterogeneity of cancers can contribute to tumor aggressiveness, invasion, and resistance to therapy. Fluorescence imaging occupies a unique niche to investigate tumor heterogeneity due to its high resolution and molecular specificity. Here, heterogeneous populations are identified and quantified by combined optical metabolic imaging and subpopulation analysis (OMI-SPA). OMI probes the fluorescence intensities and lifetimes of metabolic enzymes in cells to provide images of cellular metabolism, and SPA models cell populations as mixed Gaussian distributions to identify cell subpopulations. In this study, OMI-SPA is characterized by simulation experiments and validated with cell experiments. To generate heterogeneous populations, two breast cancer cell lines, SKBr3 and MDA-MB-231, were co-cultured at varying proportions. OMI-SPA correctly identifies two populations with minimal mean and proportion error using the optical redox ratio (fluorescence intensity of NAD(P)H divided by the intensity of FAD), mean NAD(P)H fluorescence lifetime, and OMI index. Simulation experiments characterized the relationships between sample size, data standard deviation, and subpopulation mean separation distance required for OMI-SPA to identify subpopulations.
Keywords: (000.5490) Probability theory, stochastic processes, and statistics; (100.2960) Image analysis; (170.1530) Cell analysis; (170.2520) Fluorescence microscopy; (170.6920) Time-resolved imaging; (180.4315) Nonlinear microscopy.