False Positives in Artificial Intelligence Prioritization Software for Intracranial Hemorrhage Identification in the Postoperative Period: A Report of Two Cases

Abstract

The implementation of artificial intelligence (AI) in radiology has shown significant promise in the identification of acute intracranial hemorrhages (ICHs). However, it is crucial to recognize that AI systems may produce false-positive results, especially in the postoperative period. Here, we present two cases where AI prioritization software erroneously identified an acute ICH on a postoperative non-contrast CT. These cases highlight the need for a more careful radiology review of AI-flagged images in postoperative patients to avoid further unnecessary imaging and unwarranted concerns from radiologists, clinicians, and patients.