Bias in the machines

Current challenges of implementing artificial intelligence in medical imaging

“…This has been shown in a commercial algorithm that is widely used to guide healthcare decisions in the US where it was found to be discriminative against Black patients [38]. In fact, according to a global survey, model fairness is the most ubiquitous principle in developing AI systems [39] to avoid unfair outcomes based on race and socioeconomic class [40][41][42]. Furthermore, AI models were found to have poor generalisability [34].…”

The Ethics of Artificial Intelligence in Pathology and Laboratory Medicine: Principles and Practice

“…One reason is that historical medical data may be less representative for minorities or socially marginated groups. 60 -63 Using these models for clinical decision support may further increase disparity. For example, an algorithm-based screening tool that was widely used to alert primary care doctors to high-risk patients for resource allocation was found to systematically discriminate against Black patients.…”