The discovery of detailed structures of spatial organelles within a single cell obtained by state-of-the-art molecular imaging technology has provided essential biological information for gaining insights into the study of complex human diseases. In particular, such information is helpful for cancer modeling and simulation. This paper presents a novel concept for characterizing the intracellular space of cancer and normal cells using the mathematical principle of power laws applied to a fuzzy partition functional for cluster validity. Experimental results and comparison with image texture analysis suggest the promising application of the proposed method.