Bacterial infections ranging from heart valve infections (endocarditis) to bone infections (osteomyelitis) are a growing problem in health care today and cause significant morbidity and mortality. The unmet clinical need is our inability to diagnose bacterial infections at an early stage in a specific and sensitive manner. At present, heart valve infections, bone infections and abscesses are diagnosed on the basis of clinical judgment augmented by several non-specific imaging techniques (e.g., CT, MRI, white blood cell scans) and cultures. . However, these approaches are either indirect in the case of imaging or take several days in the case of cultures. The central objective of this proposal is to develop the next generation of maltodextrin based PET contrast agents that can image and diagnose early stage gram positive and gram negative infections. We have recently demonstrated that a new optimized maltodextrin targeting ligand termed thiomalto-triose that can image gram positive bacterial infections in vivo. The following specific aims are proposed: Specific Aim 1: Synthesis and Characterization of 18F thiomalto-triose based bacterial targeting ligands Specific Aim 2: Evaluation of lead compounds for the detection of gram negative infections in an animal model of bacterial osteomyelitis. Specific Aim 3: Evaluation of lead compounds for the detection of gram positive infections in an animal model of bacterial endocarditis