Rapid diagnosis of malaria by chemometric peak-free LIBS of trace biometals in blood

Abstract

Background Laser Induced Breakdown Spectroscopy (LIBS) trace atomic species of diseased biofluids are very subtle (peak-free) in complex spectra. Trace analysis requires a considerable push in analytical strategy. Enabling LIBS with chemometrics can help identify, extract, analyze and interpret the species’ spectral signatures to give an insight on the biophysiological status of the bodies from which the biofluids originate. Methods We report on the trace quantitative performance of peak-free LIBS enabled by chemometrics modeling using principal components analysis (PCA) for direct artificial neural network (ANN)–based analysis of Cu, Zn, Fe and Mg in plasmodium falciparum-infected blood in the context of rapid spectral diagnosis of malaria utilizing the biometals as the disease biomarkers. Only one standard is required in this method - to delineate the analyte spectral regions (feature selection) and to test for accuracy. Results Based on the alteration of the biometal levels and their multivariate and correlational patterns in cultured blood, peripheral finger blood drops dried directly on Nucleopore membrane filters was accurately discriminated as either malaria-infected or healthy utilizing principal components analysis (PCA) modelling. Further the morphological evolution of plasmodium was accurately predicted using spectral features of the biometals wherein high negative correlations between Fe (-0.775) and Zn (-0.881) and high positive correlations between Cu (0.892) and Mg (0.805) with parasitemia was observed. During the first 96 hours of malaria infection Cu increases profoundly (from 328 to 1,999 ppb) while Fe, Zn and Mg decrease (from 1,206 to 674 ppb), (from 1,523 to 499 ppb) and (from 23,880 to 19,573 ppb) respectively. Compared with healthy, plasmodium falciparum-infected blood has high Cu but low levels of Fe, Zn and Mg. Cu and Zn are highly (≥0.9) positively correlated while Fe and Cu as well as Zn and Cu are highly (≥ 0.9) negatively correlated. Conclusion Chemometric peak-free LIBS has demonstrated the potential for direct rapid malaria diagnostics in blood based on the levels, alterations and multivariate associations of analyzed trace biometals which are used as biomarkers of the disease.