Stroke exacts a heavy financial and economic burden, is a leading cause of death, and the leading cause of long term disability in those who survive. The penumbra surrounds the ischemic core of the stroke lesion and is comprised of metabolically stressed cells that may become part of the growing infarct or can potentially recover. Preserving the penumbra to promote functional recovery is a central goal in stroke research. Using synchrotron-based X-ray fluorescence imaging the elemental changes in brain tissue following ischemic stroke can be visualized. We have employed a Gaussian mixture model to cluster tissue areas based on their elemental and metabolic characteristics. The method separates the core of the infarct from healthy tissue, and also demarcates discrete regions encircling the infarct. These regions of interest can be combined with elemental and metabolic data, as well as with conventional histology. The cell populations defined by clustering provide a reproducible means of visualizing the size and extent of the penumbra at the level of the single cell and provide a critically-needed tool to track changes in metabolic status and penumbra size.