Researchers at the Beckman Institute developed a framework for understanding errors that can arise due to algorithmic bias in computed imaging systems, like MRI or CT, and may contribute to patient misdiagnosis. Their work provides insight into the factors that cause these so-called hallucinations.