For decades, the promise of stroke therapies has quietly eroded between the walls of different laboratories, where inconsistent measurement has made it nearly impossible to know whether a treatment truly works or merely appears to. Researchers at USC have answered this uncertainty with an automated MRI analysis system that reads thousands of brain scans the same way every time, across six institutions and more than two thousand animals. The work, born from a federally coordinated effort to bridge preclinical research and human trials, suggests that the path from laboratory hope to clinical rea