Chaos analysis of breast masses on dynamic magnetic resonance mammography
One of the most challenging problems in biological image analysis is the quantification of the dynamical mechanism and complexity of the intracellular space. This paper investigates potential spatial chaos and complex behavior of the intracellular space of typical cancer and normal cell images whose structural details are revealed by the combination of scanning electron microscopy and focused ion beam systems. Such numerical quantifications have important implications for computer modeling and simulation of diseases. Cancer cell lines derived from a human head and neck squamous cell carcinoma (SCC-61) and normal mouse embryonic fibroblast (MEF) cells produced by focused ion beam scanning electron microscopes were used in this study. Spatial distributions of the organelles of cancer and normal cells can be analyzed at both short range and long range of the bounded dynamical system of the image space, depending on the orientations of the spatial cell. A procedure was designed for calculating the largest Lyapunov exponent, which is an indicator of the potential chaotic behavior in intracellular images. Furthermore, the sample entropy and regularity dimension were applied to measure the complexity of the intracellular images. Positive values of the largest Lyapunov exponents (LLEs) of the intracellular space of the SCC-61 were obtained in different spatial orientations for both long-range and short-range models, suggesting the chaotic behavior of the cell. The MEF has smaller positive values of LLEs in the long range than those of the SCC-61, and zero vales of the LLEs in the short range analysis, suggesting a non-chaotic behavior. The intracellular space of the SCC-61 is found to be more complex than that of the MEF. The degree of complexity measured in the spatial distribution of the intracellular space in the diagonal direction was found to be approximately twice larger than the complexity measured in the horizontal and vertical directions. Initial findings are promising for characterizing different types of cells and therefore useful for studying cancer cells in the spatial domain using state-of-the-art imaging technology. The measures of the chaotic behavior and complexity of the spatial cell will help computational biologists gain insights into identifying associations between the oscillation patterns and spatial parameters of cells, and appropriate model for simulating cancer cell signaling networks for cancer treatment and new drug discovery.