Application Of Raman Microscopy For The Diagnosis Of The Chronic Lymphocytic Leukemia (Cll)

Introduction/ Background In hematology, actual diagnosis of B chronic lymphocyte-leukemia (CLL) is based on the microscopic analysis of cell morphology from patient blood smear. However, new photonic technologies appear promising to facilitate and improve the early diagnosis, prognostic and monitoring of personalized therapy. The development of automated diagnostic approaches could assist clinicians in improving the efficiency and quality of health services, but also reduce medical costs. Aims The M3S project aims at improving the diagnosis and prognosis of the CLL pathology by developing a multimodal microscopy platform, including Raman spectrometry, dedicated to the automatic analysis of lymphocytes. Methods Blood smears were prepared on glass slides commonly used in pathology laboratories for microscopy. Two types of sample per patient were prepared: a conventional blood smear and a deposit of “pure” lymphocyte subtypes (i.e. normal B, CLL B, T and NK), sorted out in flow cytometry by using the negative double labeling technique. The second sample is used for the construction of a database of spectral markers specific of these different cell types. The preparations were analyzed with the multimodal machine which combines i) a Raman micro-spectrometer, equipped with a 532nm diode laser excitation source; ii) a microscope equipped with 40x and 150x lenses and a high precision xyz motorized stage for scanning the blood smear, and localizing x-y coordinates of representative series (~100 for each patient) of lymphocyte cells before registering three Raman spectra; these cells of interest being previously localized by an original method based on the morphology analysis. After the Raman acquisitions, the conventional blood smears were submitted to immunolabelling using specific antibodies. For the establishment of the Raman classifiers, this post-acquisition treatment was used as reference to distinguish the different lymphocyte sub-populations. Raman data were then analyzed using chemometric processing and supervised statistical classifiers in order to construct a spectral library of markers highly specific of the lymphocyte type and status (normal or pathological). Results Currently, a total of 60 patients (CLL and healthy) were included in the study. Various classification methods such as LDA (Linear Discriminant Analysis), PLS-DA (Partial Least Square Discriminant Analysis), RF (Random Forest) and SVM (Support Vector Machine), were tested in the purpose to distinguish tumoral B lymphocytes from other cell types. These classification algorithms were combined with feature selection approaches. The best performances were around 70% of correct identification when a three-class model (B-CLL vs B-normal vs T and NK lymphocytes) was considered, and 80% in case of a two-class model (B-CLL vs B-normal lymphocytes). These encouraging results demonstrate the potential of Raman micro-spectroscopy coupled to supervised classification algorithms for leukemic cell classification. The approach can find interest more generally in the field of cyto-hematology. Further developments will concern the integration of additional modality such as Quantitative Phase Imaging on one hand to speed the exploration process of cells of interest to be probed, and on the other hand to extract additional characteristics likely to be informative for CLL diagnosis. In addition, the identification of prognostic markers will be investigated by confronting the photonic data to clinical patient information.

Analytical capabilities of chemiresistive microsensor arrays in a simulated Martian atmosphere

The problems of adulteration and mislabeling are very common in the food industry. Laser-induced breakdown spectroscopy (LIBS) coupled with chemometric methods has many intrinsic advantages on adulteration analysis of various materials. In this work, several chemometric algorithms, i.e., principal component analysis (PCA), decision tree (DT), random forest (RF), partial least squares discriminant analysis (PLS-DA), linear discriminant analysis (LDA), and support vector machine (SVM), were carried out assisted by LIBS technology to study the classification performances on rice geographic origins. A series of samples, including 20 kinds of rice samples from different geographic origins, was detected using LIBS with no pretreatment processes. For data analysis, PCA was employed to reduce the input variables, and to reduce the collinearity of LIBS spectral results as well. The results showed the classification accuracies of the mentioned chemometric algorithms of DT, RF, PLS-DA, LDA, and SVM with 89 input variables of 86.80%, 96.30%, 96.80%, 98.60%, and 99.20%, respectively. At the same time, the operation times of these algorithms were 3.81 s, 54.64 s, 3.63 s, 2.09 s, and 531.01 s, respectively. On the other hand, 30 principal components of input variables were also tested under the same conditions. The classification accuracies for the above algorithms were 81.60%, 98.00%, 95.70%, 98.40%, and 99.20%, respectively. The operation times were 2.01 s, 4.88 s, 3.67 s, 0.36 s, and 308.55 s, respectively. In addition, the five-fold cross-validation classification accuracies with 30 input variables for DT, RF, PLS-DA, LDA, and SVM were 83.75%, 97.95%, 94.75%, 98.35%, and 99.25%, respectively. As a result, LDA was demonstrated to be the best and most efficient tool for rice geographic origin classification assisted by LIBS with high accuracy and analytical speed, which has great potential for rapid identification of adulterated products in agriculture without use of any chemical reagent.

In the present study, boosting has been combined with partial least‐squares discriminant analysis (PLS‐DA) to develop a new pattern recognition method called boosting partial least‐squares discriminant analysis (BPLS‐DA). BPLS‐DA is implemented by firstly constructing a series of PLS‐DA models on the various weighted versions of the original calibration set and then combining the predictions from the constructed PLS‐DA models to obtain the integrative results by weighted majority vote. Coupled with near infrared (NIR) spectroscopy, BPLS‐DA has been applied to discriminate different kinds of tea varieties. As comparisons to BPLS‐DA, the conventional principal component analysis, linear discriminant analysis (LDA), and PLS‐DA have also been investigated. Experimental results have shown that the inter‐variety difference can be accurately and rapidly distinguished via NIR spectroscopy coupled with BPLS‐DA. Moreover, the introduction of boosting drastically enhances the performance of an individual PLS‐DA, and BPLS‐DA is a well‐performed pattern recognition technique superior to LDA. Copyright © 2012 John Wiley & Sons, Ltd.

Large‐scale dieback of ash trees (Fraxinus spp.) caused by the fungus Hymenoscyphus fraxineus is posing an immense threat to forest health in Europe, requiring effective monitoring at large scales. In this study, a pipeline was created to find ash trees and classify dieback severity using high‐resolution hyperspectral imagery of individual tree crowns (ITCs). Hyperspectral data were collected in four forest sites near Cambridge, UK, where 422 ITCs were manually delineated and labelled using field‐measurements of species and dieback severity (for ash trees). Four algorithms, namely linear discriminant analysis (LDA), principal components analysis coupled with LDA (PCA‐LDA), partial least squares discriminant analysis (PLS‐DA) and random forest (RF), were used to build classification models for species and dieback severity classification. The effect of dark‐pixel filtering on classification accuracy was evaluated. The best performing models were then coupled with automatic ITC segmentation to map species and ash dieback distribution over 16.8 hectares of woodland. We calculated and partitioned the coefficient of variation (CV) of the reflected ash spectra to find variable wavebands associated with dieback. PLS‐DA and LDA were most accurate for classifying ITC species identifies (overall accuracy >90%), whereas RF was most accurate for classifying ash dieback severity (overall accuracy 77%). Dark pixel filtering further increased the accuracy of species classification (+6%), but not disease classification. The reflectances of narrow blue (415 nm), red‐edge (680 nm) and NIR (760 nm) bands had high CV across disease classes and should be included if multispectral imagery were to be used to monitor ash dieback. The study demonstrates the possibility of using remote sensing to forward epidemiological research by monitoring forest pathogens in landscape scales, which would allow temperate forest managers to control pathogen outbreaks, assess associated impacts and restore affected forests much more effectively.

BackgroundSmall sample sizes combined with multiple correlated endpoints pose a major challenge in the statistical analysis of preclinical neurotrauma studies. The standard approach of applying univariate tests on individual response variables has the advantage of simplicity of interpretation, but it fails to account for the covariance/correlation in the data. In contrast, multivariate statistical techniques might more adequately capture the multi-dimensional pathophysiological pattern of neurotrauma and therefore provide increased sensitivity to detect treatment effects.ResultsWe systematically evaluated the performance of univariate ANOVA, Welch’s ANOVA and linear mixed effects models versus the multivariate techniques, ANOVA on principal component scores and MANOVA tests by manipulating factors such as sample and effect size, normality and homogeneity of variance in computer simulations. Linear mixed effects models demonstrated the highest power when variance between groups was equal or variance ratio was 1:2. In contrast, Welch’s ANOVA outperformed the remaining methods with extreme variance heterogeneity. However, power only reached acceptable levels of 80% in the case of large simulated effect sizes and at least 20 measurements per group or moderate effects with at least 40 replicates per group. In addition, we evaluated the capacity of the ordination techniques, principal component analysis (PCA), redundancy analysis (RDA), linear discriminant analysis (LDA), and partial least squares discriminant analysis (PLS-DA) to capture patterns of treatment effects without formal hypothesis testing. While LDA suffered from a high false positive rate due to multicollinearity, PCA, RDA, and PLS-DA were robust and PLS-DA outperformed PCA and RDA in capturing a true treatment effect pattern.ConclusionsMultivariate tests do not provide an appreciable increase in power compared to univariate techniques to detect group differences in preclinical studies. However, PLS-DA seems to be a useful ordination technique to explore treatment effect patterns without formal hypothesis testing.

Characterizing and classifying molecular variations within biological samples are critical for determining the fundamental mechanisms of biological processes. Toward these ends, time‐of‐flight secondary ion mass spectrometry (ToF‐SIMS) was used to examine increasingly complex samples of biological relevance. The large, multivariate datasets were analyzed using five common statistical and chemometric techniques: principal component analysis (PCA), linear discriminant analysis (LDA), partial least‐squares discriminant analysis (PLSDA), soft independent modeling of class analogy (SIMCA), and decision‐tree analysis by recursive partitioning. PCA was found to provide insight into both the relative groupings of samples and the molecular basis for those groupings. For monosaccharide, pure protein, and complex protein mixture samples, LDA, PLSDA, and SIMCA all produced excellent classification. For mouse embryo tissues, however, SIMCA did not classify samples as accurately. The decision‐tree analysis was the least successful for all tested samples, providing neither as accurate a classification nor chemical insight. Based on these results we conclude that as the complexity of samples increases, so must the sophistication of the multivariate technique used to classify the samples. PCA is a preferred first step for understanding ToF‐SIMS data that can be followed by either LDA or PLSDA for effective classification. This study demonstrates the strength of the combination of ToF‐SIMS and multivariate analysis to classify increasingly complex biological samples. Applying these techniques to information‐rich mass spectral data opens the possibilities for new applications including classification of subtly different biological samples that may provide insights into cellular processes, disease progress, and disease diagnosis. Copyright © 2008 John Wiley & Sons, Ltd.

Detection and Follow Up of BCR/ABL Translocation by Cromogen In Situ Hibridization with ABL Centromeric and Telomeric (T/C) Probe Pair. Molecular, Accurate and Inexpensive Assay.

Hybrid Random Walk-Linear Discriminant Analysis Method for Unwrapping Quantitative Phase Images of Biological Samples

Classification of geographical origins and prediction of δ13C and δ15N values of lamb meat by near infrared reflectance spectroscopy

Moving window partial least-squares (MWPLS) regression was coupled with near infrared (NIR) spectra as an interval selection method to improve the performance of partial least squares discriminant analysis (PLSDA) models. This method was applied to the identification of artificial bezoar, natural bezoar and artificial bezoar in natural bezoar and compared with some traditional pattern recognition methods, such as principal component analysis (PCA), linear discriminant analysis (LDA) and PLSDA. The introduction of MWPLS enhanced the performance of PLSDA model. The results obtained showed that moving window partial least-squares discriminant analysis (MWPLSDA) can extract wavelength intervals with useful information and build simple yet effective classification models that can significantly improve the classification accuracy. Then MWPLSDA was used to identify natural bezoar by geographical origin; a promising result was achieved. The work showed that MWPLSDA could be a promising method for quality analysis and discrimination of chinese medical herbs according to geographical origin.

Automatic whole blood cell analysis from blood smear using label-free multi-modal imaging with deep neural networks

Classification of sugar using digital imaging and pattern recognition techniques

In this work, an investigation using UHPLC-Q-Orbitrap-MS and multivariate statistics was conducted to obtain the lipid fingerprint of Camembert cheese and to explore its correlated variation with respect to X-ray irradiation treatment. A total of 479 lipids, categorized into 16 different lipid subclasses, were measured. Furthermore, the identification of oxidized lipids was carried out to better understand the possible phenomena of lipid oxidation related to this technological process. The results confirm that the lipidomic approach adopted is effective in implementing the knowledge of the effects of X-ray irradiation on food and evaluating its safety aspects. Furthermore, Partial Least Squares-Discriminant Analysis (PLS-DA) and Linear Discriminant Analysis (LDA) were applied showing high discriminating ability with excellent values of accuracy, specificity and sensitivity. Through the PLS-DA and LDA models, it was possible to select 40 and 24 lipids, respectively, including 3 ceramides (Cer), 1 hexosyl ceramide (HexCer), 1 lysophosphatidylcholine (LPC), 1 lysophosphatidylethanolamine (LPE), 3 phosphatidic acids (PA), 4 phosphatidylcholines (PC), 10 phosphatidylethanolamines (PE), 5 phosphatidylinositols (PI), 2 phosphatidylserines (PS), 3 diacylglycerols (DG) and 9 oxidized triacylglycerols (OxTG) as potential markers of treatment useful in food safety control plans.

ABSTRACTA rapid and sensitive method for classification of virgin and recycled expanded polystyrene (EPS) food containers was developed using Fourier transform infrared spectroscopy (FTIR) and chemometrics. This method includes preparing a transparent film by dissolution, examining by FTIR and developing classification models. The degradation of EPS containers occurring during the recycling process was reflected by the carbonyl region of the infrared spectrum which was used as variables for multivariate data analysis. PCA was used to reduce the data dimension and view the sample similarities. Soft independent modelling of class analogy (SIMCA), partial least squares-discrimination analysis (PLS-DA) and linear discrimination analysis (LDA) were applied to construct three classification models. The best discrimination results were obtained by an LDA model, with all samples correctly classified. PLS-DA and SIMCA could not classify the recycled EPS samples with low levels of adulteration. When applying this method to commercially available EPS containers, about 45% of samples were shown to contain recycled polystyrene resins. It is concluded that the carbonyl region of the infrared spectra coupled with chemometrics could be a powerful tool for the classification of virgin and recycled EPS food containers.