Contamination of tissue samples can mislead AI models, preventing them from making accurate diagnoses in real-world situations.
Human pathologists undergo rigorous training to identify instances where tissue samples from one patient are accidentally placed on microscope slides meant for another patient, a mistake referred to as tissue contamination. However, this type of contamination poses a significant challenge for artificial intelligence (AI) models, which are typically developed in clean, controlled settings, according to a recent study by Northwestern Medicine.
“We train AIs to tell ‘A’ versus ‘B’ in a very clean, artificial environment, but, in real life, the AI will see a variety of materials that it hasn’t trained on. When it does, mistakes can happen,” said corresponding author Dr. Jeffery Goldstein, director of perinatal pathology and an assistant professor of perinatal pathology and autopsy at DOI: 10.1016/j.modpat.2024.100422
The study was funded by the National Institute of Biomedical Imaging and Bioengineering, the National Center for Advancing Translational Sciences (NCATS), the Walder Foundation Fund to Retain Clinician Scientists, and Department of Health and Human Services.