Abstract

The study of biological phenomena and physiological parameters relies on precise measurements. However, many mistakenly believe that a measurement is just a simple numerical value representing the ratio between the quantity being measured and its unit. This is far from true! Measurement is a complex process influenced not only by potential errors but also by multiple factors that make the “true” value of a biomedical parameter inherently uncertain. Recognizing this uncertainty is essential for making informed medical and scientific decisions, especially when relying on artificial intelligence (AI) for diagnostics and treatment recommendations. The goal of this seminar is to introduce methods and approaches aligned with the UNI CEI ENV 13005 standard: “Guide to the Expression of Measurement Uncertainty.” This framework enables the quantification of uncertainty in biomedical measurements, ensuring a clearer understanding of the reliability of medical data—an essential step when integrating AI-driven algorithms in healthcare. Quantifying uncertainty is crucial for comparing diagnostic results, assessing data quality, and ultimately making well-informed healthcare decisions. When AI algorithms process medical data, any inaccuracy in measurement can lead to misleading predictions, incorrect diagnoses, or flawed treatment plans.