Yuri Markushin

Tag-femtosecond laser-induced breakdown spectroscopy for the sensitive detection of cancer antigen 125 in blood plasma

Successful treatment of cancers requires detecting early signs of the disease. One promising way to approach this is to develop minimally invasive tests for the sensitive and specific detection of biomarkers in blood. Irrespective of the detection approach one uses, this remains a challenging task because biomarkers are typically present in low concentrations and there are signals that interfere strongly with prevailing compounds of human fluids. In this paper, we show that elemental encoded particle assay coupled with femtosecond laser-induced breakdown spectroscopy for simultaneous multi-elemental analysis can significantly improve biomarker detectability. An estimated near single molecule per particle efficiency of this method leads to sensitive detection of ovarian cancer biomarker CA125 in human blood plasma. This work opens new ways for earlier detection of cancers and for multiplex assay developments in various analytical applications from proteomics, genomics, and neurology fields.

Determination of protein hydrogen composition by laser-induced breakdown spectroscopy

We report on the quantitative identification of the hydrogen composition of high molecular weight proteins in a heavy water solution by laser-induced breakdown spectroscopy.

Elemental analysis of laser induced breakdown spectroscopy aided by an empirical spectral database

Laser induced breakdown spectroscopy (LIBS) is commonly used to identify elemental compositions of various samples. To facilitate this task, we propose the use of an elemental spectral library for single-pulsed, nanosecond LIBS in the spectral range 198-968 nm. This spectroscopic library is generated by measuring optical emissions from plasmas of 40 pure elements. To demonstrate the usefulness of the proposed database, we measure and analyze the LIBS spectra of pure iron and of ethanol and show that we identify these samples with a high degree of certainty.

Automatic Classification of Laser-Induced Breakdown Spectroscopy (LIBS) Data of Protein Biomarker Solutions

We perform multi-class classification of laser-induced breakdown spectroscopy data of four commercial samples of proteins diluted in phosphate-buffered saline solution at different concentrations: bovine serum albumin, osteopontin, leptin, and insulin-like growth factor II. We achieve this by using principal component analysis as a method for dimensionality reduction. In addition, we apply several different classification algorithms (K-nearest neighbor, classification and regression trees, neural networks, support vector machines, adaptive local hyperplane, and linear discriminant classifiers) to perform multi-class classification. We achieve classification accuracies above 98% by using the linear classifier with 21–31 principal components. We obtain the best detection performance for neural networks, support vector machines, and adaptive local hyperplanes for a range of the number of principal components with no significant differences in performance except for that of the linear classifier. With the optimal number of principal components, a simplistic K-nearest classifier still provided acceptable results. Our proposed approach demonstrates that highly accurate automatic classification of complex protein samples from laser-induced breakdown spectroscopy data can be successfully achieved using principal component analysis with a sufficiently large number of extracted features, followed by a wrapper technique to determine the optimal number of principal components.

LIBS-based multi-element coded assay for ovarian cancer application

We report on a new application of laser induced breakdown spectroscopy (LIBS) for the diagnosis of diseases such as ovarian cancer. We perform detection of ovarian cancer biomarker CA 125 based on LIBS measurements. Immunoconjugated Silicon particles are incubated with the affinity agarose beads carrying CA125 molecules. In the competitive affinity method Si particles carrying IgG molecules are pre-incubated with CA125. This pre-incubation decreases the numbers of free IgG molecules available for consequent interaction with the affinity beads. Thus less Si particles are attached to the agarose beads and consequently smaller Si peak area is measured by LIBS. We demonstrate a limit-ofdetection about 30 ppb for model protein avidin. We use two-element coded micro-particles to yield spectroscopic emission code using LIBS. We show that LIBS-based data collecting technique provides methodology for identification of biomarkers and cost-effective device for future clinical applications.

A Family of Chisini Mean Based Jensen-Shannon Divergence Kernels

Jensen-Shannon divergence is an effective method for measuring the distance between two probability distributions. When the difference between these two distributions is subtle, Jensen-Shannon divergence does not provide adequate separation to draw distinctions from subtly different distributions. We extend Jensen-Shannon divergence by reformulating it using alternate operators that provide different properties concerning robustness. Furthermore, we prove a number of important properties for this extension: the lower limits of its range, and its relationship to Shannon Entropy and Kullback-Leibler divergence. Finally, we propose a family of new kernels, based on Chisini mean Jensen-Shannon divergence, and demonstrate its utility in providing better SVM classification accuracy over RBF kernels for amino acid spectra. Because spectral methods capture phenomenon at subatomic levels, differences between complex compounds can often be subtle. While the impetus behind this work began with spectral data, the methods are generally applicable to domains where subtle differences are important.

Performance of multilayer perceptrons for classification of LIBS protein spectra

We investigate performance of neural networks for classification of laser-induced breakdown spectroscopic data of four proteins: Bovine Serum Albumin, Osteopontin, Leptin and Insulin-like Growth Factor II. We utilize principal component analysis algorithm for feature extraction and multilayer perceptrons algorithms with one and two hidden layers. We employ leave-one-out procedure for classifier evaluation. Our experimental results indicate that methods with linear convergence can provide classification accuracy superior to methods with quadratic convergence.

Statistical Analysis for Automatic Identification of Ovarian Cancer Protein-Biomarkers Based on Fast Fourier Transform Infrared Spectroscopy

Marcano A.1,2,3, D. Pokrajac1,4, A. Lazarevic1, M. Smith3, Y. Markushin1,2 and N. Melikechi1,2,3 1Center for Research and Education in Optical Sciences and Applications, 2Center for Applied Optics for Space Science, 3Department of Physics and Pre-Engineering 4Department of Computer and Information Sciences, Delaware State University, 1200 North Dupont Highway, Dover, DE 19901 United States of America

Fourier transform infrared spectroscopy of deuterated proteins

We report on Fourier transform spectra of deuterated proteins: Bovine Serum Albumin, Leptin, Insulin-like Growth Factor II, monoclonal antibody to ovarian cancer antigen CA125 and Osteopontin. The spectra exhibit changes in the relative amplitude and spectral width of certain peaks. New peaks not present in the non-deuterated sample are also observed. Ways for improving the deuteration of proteins by varying the temperature and dilution time are discussed. We propose the use of deuterated proteins to increase the sensitivity of immunoassays aimed for early diagnostic of diseases most notably cancer.