Targeting and quantifying metabolic changes non-invasively is a powerful approach to facilitate diagnosis and evaluate therapeutic response. Cellular metabolism involves a vital network of pathways for homeostasis, growth, and survival and can shift from one nutrient pathway to another based on the extent of perfusion available to the cells. Interest in cellular metabolism and tissue vasculature continues to expand across a broad range of disciplines including neuroscience, cardiovascular biology, and the field of cancer research. Though there are many bench-top microscopes and metabolic tools available to provide exquisite resolution and contrast for metabolic or vascular imaging, these systems require extensive training and often have fields of view (FOV), resolution, and wavelengths that fit only the most common use cases. Further, they require researchers to transport animals to specialized facilities, and this limits access to longitudinal imaging. Additionally, there are surprisingly few biomedical imaging technologies available to image both the global landscape and local spatial variations of metabolic and vascular hallmarks in vivo. We propose to develop an optical imaging platform the Capillary-Cell or CapCell to permit studies of metabolic reprogramming and heterogeneity across the laboratory to clinical continuum. This technology will report on the major axes of metabolism, blood vessel architecture and morphology of different biological systems including organoids and xenograft mouse models. Our work will lead to the establishment of predictive biomarkers to support drug development, inform on drug choices and evaluate the efficacy of drugs in bench research and in patients. The technology will be portable and turnkey and therefore can be placed in individual labs instead of a central dedicated facility. This is essential to putting new biomarker capabilities directly into the hands of laboratory scientists. The clinical translatability of the CapCell will focus on breast cancer therapies. The CapCell will inform the selection of compounds for personalized management of cancer patients in adaptive clinical trials and ultimately those in a standard clinical setting. Further, it will enable the identification of successful drugs early in their development, thereby accelerating market approval for candidate therapies. Lastly, this technology will be instrumental in understanding metabolic heterogeneity during primary cancer formation and invasion and its modulation by oncogenic driver mutations and their inhibition. The biological models will include human organoids and mouse models developed at UCSF. Patient-derived organoids (PDO) will serve as a bridge between patients and mouse models – they can be created from patient samples, and they can then be engrafted into mice, and Patient-derived xenografts (PDX) models can be used to identify dynamic biomarkers associated with the risk of recurrence.