Is the poison/explosive/biomarker really there?

“Estimation of the Limit of Detection Using Information Theory Measures” addresses the problem of the Limit of Detection (LOD) in analytical methods or measurement instruments. The LOD specifies the smallest quantity at which an analyte can be detected or distinguished from a measurement performed in the absence of the analyte. This means it is a limiting factor in the chemical detection systems – which are used in such wide-ranging applications such as disease diagnosis, environmental monitoring, poisons detection and security – where false positives or negatives can be disastrous.

“The LOD given in the specifications of a system is very significant, as it may have legal implications or play a role in market regulations,” explains IBEC’s Santiago Marco, who worked with ex-IBEC postdoc Jordi Fonollosa, now at the University of California San Diego, and other colleagues in the U.S. on the paper. “For example, in 2006 the US Environmental Protection Agency revised the maximum admissible contamination level of arsenic in drinking water from 50 nmol per mol to 10 nmol per mol. As a consequence, based on their LOD, some of the existing techniques to measure arsenic concentration in water were completely invalidated.”

With this in mind, the researchers propose a new way of defining the LOD in which, unlike in existing probabilistic methods, the relevant magnitude becomes the amount of information (in bits) the measurement provides about the presence or absence of the analyte. This more principled approach easily shows that the noise type is an important factor on the information that may be extracted from the measurement.

“The limit of detection is a fundamental concept in measurement science,” says Santiago, who was also recently appointed Associate Editor of the IEEE Sensors Journal. “We hope that regulatory bodies like the International Union of Pure and Applied Chemistry and the International Organization for Standarization might consider our definition in future revisions of the terminology.”

The journal cover shows a scheme of the principle behind the new proposed methodology.


Source article: Jordi Fonollosa, Alexander Vergara, Ramón Huerta, Santiago Marco (2013). Estimation of the limit of detection using information theory measures. Analytica Chimica Acta, 810, 1–9