Since the advent of magnetic resonance imaging (MRI) in the 1970s, the field of medical imaging has witnessed remarkable progress, allowing unprecedented examination of both structural and functional aspects of the human body. Among these developments, functional MRI (fMRI) has emerged as a powerful modality for probing neural activity and deciphering the intricate patterns of connectivity that underlie cognition, perception, and motor control. Resting-state functional MRI (rsfMRI), in particular, provides a non-invasive means of observing the brain’s intrinsic functional organization without the need for explicit behavioral tasks, making it especially valuable for investigating complex pathologies such as ischemic stroke, where network-level disruptions can critically influence patient prognosis. This thesis establishes and validates a rigorous cross-species functional connectivity mapping framework designed to enhance the translational relevance of preclinical stroke models for clinical neuroscience. Focusing on canine models of acute ischemic stroke, it leverages advanced MRI acquisition protocols, comprehensive preprocessing pipelines, and sophisticated computational methods—including manifold alignment algorithms, nonlinear registration techniques, and graph-theoretic analyses—to characterize stroke-induced alterations in large-scale brain networks. Critically, this research evaluates how novel hemodynamic and oxygenation-enhancing interventions, specifically NEH and Sanguinate, modulate these disrupted networks. By quantitatively assessing network reorganization and functional recovery patterns in canines treated with these agents, the framework identifies conserved features of connectivity that can be mapped onto human stroke data. Building on this cross-species alignment, the thesis employs predictive modeling strategies to infer potential therapeutic outcomes in human stroke patients. Machine learning tools are used to integrate animal-derived biomarkers, connectivity metrics, and inferred network topologies into predictive models that estimate patient-specific responses to analogous interventions. This approach capitalizes on the biophysical parallels between canine and human cerebrovascular systems, thereby reducing uncertainties associated with direct extrapolation and improving the reliability of translational insights. The outcomes of this research reinforce the notion that intrinsic network dynamics, captured via rsfMRI, offer crucial information about tissue viability, metabolic demands, and the capacity for functional reorganization following ischemic injury. Moreover, by methodically bridging the gap between preclinical and clinical domains, the thesis demonstrates how cross-species connectivity mapping can inform precision medicine approaches in stroke care. Taken together, these findings not only advance fundamental knowledge of stroke-induced network perturbations but also pave the way for more targeted, evidence-based clinical interventions and the refinement of therapeutic strategies aimed at improving patient outcomes in neurological rehabilitation.