U.S. flag

An official website of the United States government, Department of Justice.

Establishing Sufficiency Thresholds for Assessing the Quality of Mass Spectral Data

NCJ Number
303407
Author(s)
Preshious Rearden, Ph.D.
Date Published
2021
Length
37 pages
Annotation

The methodology and findings are reported for a project with the goal of developing and testing a sufficiency standard (quality value) for mass spectral data by building a spectral library of opioids across a wide range of concentrations, contaminant interference, and substance mixtures.

Abstract

As forensic laboratories increasingly encounter novel or synthetic analog drugs, library matching or pattern recognition has become more challenging. The most widely accepted practice for mass spectral identification is the analyst’s impression of concordance with existing library examples of established composition; however, samples with adulterants or high noise-to-signal ratio pose greater difficulty for delineation of a sufficiency threshold for interpretable data. From its experimental database, the current project used statistical/mathematical methodology for pattern data to create and validate an Information Quality Model for determining analytical thresholds for sufficiency of information. The project was implemented as proposed. The quantitative reliability metric (QRM) developed in this study was used to determine the probability of accuracy for a given mass spectrum searched against the library database. Although mass spectra produced from opioids were the focus of this study, the process used could be extended to other controlled substances and compounds. The methodology can be directly applied or adapted by other forensic laboratories. In addition, both the results and methodology of this project should have a direct extension to other forensic disciplines that use mass spectral data, such as Toxicology and Trace Analysis. The QRM can be applied to spectral database searching and with any query of a collection of reference data. Thus, it provides a uniform statistical probability of the reliability of search results and can be applied to database searching in other industries. 9 tables, 7 figures, and 10 references