Determining the Sources of Measurement Uncertainty in Environmental Cell-Based Biosensing

Measurement uncertainty has become an important concept in quantitative chemical analysis that unifies many previously disparate strands of information on data quality. When performing the evaluation, direct application of the guide to the expression of uncertainty in measurement may not be feasible, unless the measurement uses a straightforward detection mode and follows a well-established model at a given number of parameters. In a microbial biosensor platform, however, a variety of interrelated processes occur simultaneously or consecutively during target recognition and quantification to form a dynamic and perplex network of events that end up with an output signal correlated, sometimes erroneously, only to the target analyte. In environmental applications, the complexity of the measuring system rivals the complexity of the measurand. To handle such a dynamic system, the authors propose a knowledge-management tool relying on fuzzy fault tree analysis for identifying error and uncertainty in microbial biosensor detectors with respect to: 1) sampling; 2) method of analysis (sensing); 3) instrument; and 4) data processing. Thereby, an expert system has been developed where the tree structure serves as the knowledge base and the fuzzy rules-based decision mechanism is the inference engine.