An FDA-cleared artificial intelligence-based tool had lower positive predictive values and sensitivity when used for intracranial hemorrhage detection, compared with previous studies, according to a study in the Journal of the American College of Radiology. The findings, which showed an association between lower performance and a neurosurgery history and ICH features, "further highlight the need for standardized study design to allow for rigorous and reproducible site-to-site comparison of emerging deep learning technologies," study authors wrote.