Parametric images from the tissues of living animals and, we hope, eventually human subjects gives us the opportunity to ask how we understand the amount of information such images make available. Images are formed from data obtained from a variety of imaging methods. Electron paramagnetic resonance through the use of magnetic field gradients allows acquisition of data partitioning a living sample into subvolumes. An injectable paramagnetic spin probe with relatively long lived unpaired electrons allows us to measure the local oxygen concentrations in the aqueous extracellular fluid compartment of living animals. Oxygen, crucial as the ultimate oxidative substrate of living processes, is, of course crucial for homeostasis. To better than 100 pp,, it provides the relaxation lattice to which the magnetization energy of excited magnetization states of the spin probe distributes its energy. Unlike linewidth or equivalently spin-spin relaxation rate based oxymetry, which can be confounded by the effect of the spin probe concentration itself, the spin lattice relaxation is nearly an order of magnitude less susceptible to self-contamination, making it a near absolute oximetry. The issue, however, is how independent are the image subvolumes or voxels, defined by the spatial resolution of the EPR image. Living systems demonstrate collective coordinated behavior over a variety of distance and time scales. Cells in a multicellular organism exhibit coordination through cell signaling mechanisms, many involving secretion and either diffusion or conveyance through vascular sytems (endocrine signaling) of specific signaling molecules. Nerve conduction of signals more rapid and more extensive in organism coordination. These processes obtain in tissues and organs of healthy animals. Cancers represent more chaotic growths. Cancer blood vessels demonstrate massive dysfunction. Sensory nerve conduction is typically muted in cancers and inferentially, efferent signals are similarly blunted. On the other hand, the treatment of malignancy with cytotoxins such as chemotherapy and radiation often restores an underlying scaffolding providing at least partially functional tissues. Understanding the extent of correlation/independence of image voxels and is therefore a complicated process. Developing methods by which to evaluate the covariation or the relative independence of image voxels will allow assessing the information content of parametric images such as EPR pO2 images.