Applying virtual sample generation and ensemble modeling for improving the spectral diagnosis of cancer.

Cancer diagnosis plays a key role in facilitating treatment and improving survival rates of patients. The combination of near-infrared (NIR) spectroscopy with data-driven algorithms offers a rapid and cost-effective approach for such a task. Due to the limitations of objective cases, the number of tumor samples is usually smaller, and the resulting dataset exhibit the issues of class imbalance, which has a more serious impact on the performance of diagnostic models. To deal with class imbalance and improve the sensitivity, this work investigates the feasibility of NIR spectroscopy combined with virtual sample generation (VSG) as well as ensemble strategy for developing diagnostic models. Based on preliminary experiment, several learning algorithms such as discriminant analysis (DA) and partial least square-discriminant analysis (PLS-DA) are screened out as algorithms for constructing prediction models. Three algorithms of VSG including synthetic minority oversampling technique (SMOTE), Borderline-SMOTE and adaptive synthetic sampling (ADASYN) are used for experiment. A fixed sample subset composed of 27 cancer samples and 54 normal samples are hold out as the test set. Three training sets containing 5, 10, 25 minority class samples and 54 majority class samples are used for model development. The experimental result indicates that overall, with PLS-DA algorithm, all VSG approaches can significantly improve the sensitivity of cancer diagnosis for all cases of training sets with different minority samples, but ADASYN performs the best. It reveals that the integration of NIR, PLS-DA, and ADASYN is a promising tool package for developing diagnosis methods.

Detection of glibenclamide adulterated in antidiabetic Chinese patent medicine by attenuated total reflectance -infrared spectroscopy and chemometrics.

There have been many reports of adulterated Chinese patent medicine with synthetic prescription that are claimed to be "pure natural". The present work investigates the feasibility of combining attenuated total reflectance-Mid-infrared (ATR-MIR) spectroscopy and several interval-based PLS algorithms for detecting the glibenclamide illegally adulterated in antidiabetic Chinese patent medicine (Jiangtangning). The full-spectrum PLS, four kinds of traditional interval PLS algorithms (iPLS, biPLS, siPLS and mwPLS) and a modified algorithm, i.e., a combination of mwPLS and window size optimization, named cmwPLS, were used for building calibration models. A total of 21 samples adulterated with 0-3.5% glibenclamide were prepared. The dataset was equally split into a training set and a test set for building and testing the prediction models, respectively. For those interval-based PLS, the whole wavenumber axis was divided into 20 sub-intervals. In terms of the prediction on the test set, the new cmwPLS produce the best model, followed by mwPLS. The modified algorithm can optimize automatically the window width (i.e., the number of adjacent variables used for modeling) and position. It can be concluded that cmwPLS coupled with ATR-MIR technique is a good alternative to other traditional chemical analysis for detecting the adulteration of Chinese patent medicine.

Classification of different liquid milk by near-infrared spectroscopy and ensemble modeling.

Dairy products are necessary components of a healthy diet for human and nowadays, liquid milk become very popular because of its convenience. The identification of a brand of liquid milk is of importance. In this study, near-infrared (NIR) spectroscopy is used for rapid and objective classification of different brands of liquid milk. Chemometric methods including extreme learning machine (ELM) and its ensemble version (EELM) are investigated and compared. A dataset containing 144 samples from 6 brands are collected for experiment. A model-independent filter algorithm, i.e., relief-based feature selection, was used for variable reduction. Principal component analysis (PCA) is used as a tool of exploratory analysis for visualizing the difference among liquid milk samples of different brands. All samples were divided into three subsets, i.e., the training set, validation set and test set, for constructing, optimizing and testing the model, respectively. The model developed by the EELM procedure achieved 100% of classification accuracy, indicating that NIR spectroscopy combined with variable reduction and the EELM algorithm is feasible for classifying the brands of liquid milk.

Ensemble of extreme learning machines for multivariate calibration of near-infrared spectroscopy.

Inspired by the attractive features of extreme learning machine (ELM), a simple ensemble ELM algorithm, named EELM, is proposed for multivariate calibration of near-infrared spectroscopy. Such an algorithm takes full advantage of random initialization of the weights of the hidden layer in ELM for obtaining the diversity between member models. Also, by combining a large number of member models, the stability of the final prediction can be greatly improved and the ensemble model outperforms its best member model. Compared with partial least-squares (PLS), the superiority of EELM is attributed to its inherent characteristics of high learning speed, simple structure and excellent predictive performance. Three NIR spectral datasets concerning solid samples are used to verify the proposed algorithm in terms of both the accuracy and robustness. The results confirmed the superiority of EELM to classic PLS. Also, even if the experiment is done on NIR datasets, it provides a good reference for other spectral calibration.

Application of near-infrared spectroscopy and class-modeling to antibiotic authentication.

Nowadays, counterfeit medicines have become very popular due to the extension of the Internet. Broad-spectrum antibiotics with similar effect, but different prices, provide a gold opportunity for illegal traders to counterfeit. It is found that some genuine packaging of expensive brand drugs are recycled and then used to refill other kinds of cheap antibiotic tablets. It is of great importance to establish an effective antibiotic authentication method to check whether a product with a specific claim on its label is compatible with that declaration. In the present work, the feasibility of near-infrared (NIR) spectroscopy coupled with class-modeling for antibiotics authentication, i.e., counterfeiting between different antibiotics, is investigated. A total of 591 antibiotics samples of nine classes of different dosage forms were collected. Principal component analysis (PCA) was used for exploratory analysis. An effective model-independent filter method, i.e., relief, was used for feature selection and a novel class-modeling algorithm was used to construct authentication models. Three kinds of antibiotics were used as the target classes for experiments. The results confirmed that such a scheme is feasible and can be used in the screening of fake drug.

Soil and biomass carbon re-accumulation after landslide disturbances.

In high-standing islands of the Western Pacific, typhoon-triggered landslides occasionally strip parts of the landscape of its vegetative cover and soil layer and export large amounts of biomass and soil organic carbon (OC) from land to the ocean. After such disturbances, new vegetation colonizes the landslide scars and OC starts to reaccumulate. In the subtropical mountains of Taiwan and in other parts of the world, bamboo (Bambusoideae) species may invade at a certain point in the succession of recovering landslide scars. Bamboo has a high potential for carbon sequestration because of its fast growth and dense rooting system. However, it is still largely unknown how these properties translate into soil OC re-accumulation rates after landslide disturbance. In this study, a chronosequence was established on four former landslide scars in the Central Mountain Range of Taiwan, ranging in age from 6 to 41 years post disturbance as determined by landslide mapping from remote sensing. The younger landslide scars were colonized by Miscanthus floridulus, while after approx. 15 to 20 years of succession, bamboo species (Phyllostachys) were dominating. Biomass and soil OC stocks were measured on the recovering landslide scars and compared to an undisturbed Cryptomeria japonica forest stand in the area. After initially slow re-vegetation, biomass carbon accumulated in Miscanthus stands with mean annual accretion rates of 2 ± 0.5 Mg C ha-1 yr-1. Biomass carbon continued to increase after bamboo invasion and reached ~40% of that in the reference forest site after 41 years of landslide recovery. Soil OC accumulation rates were ~2.0 Mg C ha-1 yr-1, 6 to 41 years post disturbance reaching ~64% of the level in the reference forest. Our results from this in-situ study suggest that recovering landslide scars are strong carbon sinks once an initial lag period of vegetation re-establishment is overcome.

Express detection of expired drugs based on near-infrared spectroscopy and chemometrics: A feasibility study.

Levofloxacin is a third-generation fluoroquinolone antimicrobials drug that inhibits bacterial DNA replication. Driven by huge profit, one kind of particular counterfeit, e.g., repackaged expired tablets, becomes very common especially in developing countries. The feasibility of identifying expired levofloxacin tablets by combining NIR spectroscopy with chemometrics was investigated. Five kinds of levofloxacin samples from different manufacturers were collected for experiment. Two types of expired mode were considered and a simple model-independent algorithm was used for feature selection. Principal component analysis (PCA) was used for exploratory analysis and simple discriminant analysis. Taking seventy samples as the target class, a final one-class model based on Data Driven Soft Independent Modeling by Class Analogy with abbreviation DD-SIMCA was constructed, which achieved 97% sensitivity and 100% specificity on the independent set composed of 34 unexpired and 128 expired tablets. These results confirm that the combination of NIR spectroscopy, feature selection and class-modeling is feasible for identifying the expired levofloxacin tablets. Such a method can be extended to the analysis of similar drugs.

Non-destructive identification of native egg by near-infrared spectroscopy and data driven-based class-modeling.

Eggs are very important parts of human diets worldwide. It is very common to pass feed eggs off as native ones of high commercial values in Chinese markets. One urgent and challenging work is to develop a non-destructive method for verifying the authenticity of native eggs. The present work focuses on exploring the feasibility of combining near-infrared (NIR) spectroscopy with data driven-based class-modeling (DDCM) and model-independent variable selection, i.e., joint mutual information (JMI). A total of 122 eggs of three types were collected. Principal component analysis (PCA) was utilized for exploratory analysis. The JMI algorithm selected only 20 informative variables out of 1557 original variables for class-modeling. DDCM constructed a class-model for each kind of eggs by optimizing parameters such as degrees of freedom (DoF) and the number of principal components (NPC). All class-models and the decision rules were validated on the corresponding test sets. In short, these models achieved an acceptable performance and are also more consistent with actual needs than classification models. The results show that NIR spectroscopy combined with class-modeling is a potential tool for detecting the authenticity of a specific kind of native eggs.

Random subspace-based ensemble modeling for near-infrared spectral diagnosis of colorectal cancer.

The feasibility of using near-infrared (NIR) spectroscopy coupled with classifier ensemble for improving the diagnosis of colorectal cancer was explored. A total of 157 NIR spectra from the patients were recorded and partitioned into the training set and the test set. Four algorithms, i.e., Adaboost.M1, Totalboost and LPboost using decision tree as weak learners, together with random subspace method (RSM) using linear discriminant classifier (LDA) as weak learners, were used to construct diagnostic models. Some key parameters such as the size of ensemble, i.e., the number of weak learners in ensemble, and the size of each subspace in RSM, were optimized. The results indicated that, in terms of generalization ability, the RSM-based classifier outperforms all other classifiers by only 40 members with 30 features each. On the basis of 200 different training sets, model population analysis (MPA) was made. The average sensitivity and specificity of the RSM classifier were 97.4% and 95.6%, respectively. It indicates that the NIR technique combined with the RSM algorithm can serve as a potential means for automatic identification of colorectal tissues.

Quantifying several adulterants of notoginseng powder by near-infrared spectroscopy and multivariate calibration.

The authentication of traditional Chinese medicine (TCM) is critically important for public-health and economic terms. Notoginseng, a classical TCM of high economic and medical value, could be easily adulterated with Sophora flavescens powder (SFP), corn flour (CF) or other analogues of low-grade (ALG) because of their similar tastes, appearances and much lower cost. The main objective of this study was to evaluate the feasibility of applying of near-infrared (NIR) spectroscopy and multivariate calibration for identifying and quantifying several common adulterants in notoginseng powder. Two datasets were prepared for experiment. The competitive adaptive reweighted sampling (CARS) was used to select informative variables. Two different schemes were used for sample set partition. Model population analysis (MPA) was made. The results showed that, the constructed partial least squares (PLS) model using a reduced set of variables from CARS can provide superior performance to the full-spectrum PLS model. Also, the sample set partition is very of great importance. It seems that the combination of NIR spectroscopy, CARS and PLS is feasible to quantify common adulterants in notoginseng powder.

Fast discrimination of the geographical origins of notoginseng by near-infrared spectroscopy and chemometrics.

Notoginseng is a type of highly valued Traditional Chinese medicine (TCM) due to its hemostatic and cardiovascular functions. Notoginseng of Yunnan in China usually commands a premium price and is often the subject of fraudulent practices. The feasibility of combining near-infrared (NIR) spectroscopy with chemometrics was investigated to discriminate notoginseng of different geographical origins. A total of 250 samples of four different provinces in China were collected and divided equally into the training and test sets. Principal component analysis (PCA) was used for observing possible trend of grouping. Two chemometric algorithms including partial least squares-discriminant analysis (PLSDA) and soft independent modeling of class analogy (SIMCA) were used to construct the discriminant models. Standard normal variate (SNV) and first derivative were used for pre-processing spectra. On the independent test set, the PLSDA model outperforms the SIMCA model. When combining both pre-processing methods, the constructed PLSDA model achieved 100% sensitivity and 100% specificity on both the training set and the test set. It indicates that SNV+first derivative pre-processing and PLSDA algorithm can serve as the potential tool of fast discriminating the geographical origins of notoginseng.

Authenticity Detection of Black Rice by Near-Infrared Spectroscopy and Support Vector Data Description.

Black rice is an important rice species in Southeast Asia. It is a common phenomenon to pass low-priced black rice off as high-priced ones for economic benefit, especially in some remote towns. There is increasing need for the development of fast, easy-to-use, and low-cost analytical methods for authenticity detection. The feasibility to utilize near-infrared (NIR) spectroscopy and support vector data description (SVDD) for such a goal is explored. Principal component analysis (PCA) is used for exploratory analysis and feature extraction. Another two data description methods, i.e., k-nearest neighbor data description (KNNDD) and GAUSS method, are used as the reference. A total of 142 samples from three brands were collected for spectral analysis. Each time, the samples of a brand serve as the target class whereas other samples serve as the outlier class. Based on both the first two principal components (PCs) and original variables, three types of data descriptions were constructed. On average, the optimized SVDD model achieves acceptable performance, i.e., a specificity of 100% and a sensitivity of 94.2% on the independent test set with tight boundary. It indicates that SVDD combined with NIR is feasible and effective for authenticity detection of black rice.

Quantitative determination of wool in textile by near-infrared spectroscopy and multivariate models.

The wool content in textiles is a key quality index and the corresponding quantitative analysis takes an important position due to common adulterations in both raw and finished textiles. Conventional methods are maybe complicated, destructive, time-consuming, environment-unfriendly. Developing a quick, easy-to-use and green alternative method is interesting. The work focuses on exploring the feasibility of combining near-infrared (NIR) spectroscopy and several partial least squares (PLS)-based algorithms and elastic component regression (ECR) algorithms for measuring wool content in textile. A total of 108 cloth samples with wool content ranging from 0% to 100% (w/w) were collected and all the compositions are really existent in the market. The dataset was divided equally into the training and test sets for developing and validating calibration models. When using local PLS, the original spectrum axis was split into 20 sub-intervals. No obvious difference of performance can be seen for the local PLS models. The ECR model is comparable or superior to the other models due its flexibility, i.e., being transition state from PCR to PLS. It seems that ECR combined with NIR technique may be a potential method for determining wool content in textile products. In addition, it might have regulatory advantages to avoid time-consuming and environmental-unfriendly chemical analysis.

Classification and quantitation of milk powder by near-infrared spectroscopy and mutual information-based variable selection and partial least squares.

Milk is among the most popular nutrient source worldwide, which is of great interest due to its beneficial medicinal properties. The feasibility of the classification of milk powder samples with respect to their brands and the determination of protein concentration is investigated by NIR spectroscopy along with chemometrics. Two datasets were prepared for experiment. One contains 179 samples of four brands for classification and the other contains 30 samples for quantitative analysis. Principal component analysis (PCA) was used for exploratory analysis. Based on an effective model-independent variable selection method, i.e., minimal-redundancy maximal-relevance (MRMR), only 18 variables were selected to construct a partial least-square discriminant analysis (PLS-DA) model. On the test set, the PLS-DA model based on the selected variable set was compared with the full-spectrum PLS-DA model, both of which achieved 100% accuracy. In quantitative analysis, the partial least-square regression (PLSR) model constructed by the selected subset of 260 variables outperforms significantly the full-spectrum model. It seems that the combination of NIR spectroscopy, MRMR and PLS-DA or PLSR is a powerful tool for classifying different brands of milk and determining the protein content.

Influence of cigarette smoking on osteonecrosis of the femoral head (ONFH): a systematic review and meta-analysis.

BACKGROUND: Current studies demonstrate controversy regarding the relationship between cigarette smoking and osteonecrosis of the femoral head (ONFH). METHODS: We conducted a meta-analysis to evaluate the association between smoking and ONFH. Relevant articles published before September 2016 were identified by a systematic search of EMBASE and MEDLINE via Ovid. Summary odds ratios (OR) were calculated using random effects models, and study quality was assessed using a modified Newcastle-Ottawa scale. RESULTS: 102 citations were screened and 7 case-control studies were identified and included in the review. When compared with nonsmokers, current smokers had a higher risk of developing ONFH (OR 2.53; 95% confidence interval [CI] 1.68-3.79), as did former smokers (OR 1.82; 95% CI, 1.10-3.00). Within the group of current smokers, those classified as heavy smokers (with a daily number >20 cigarettes/day) demonstrated higher risks of ONFH (OR 2.03; 95% CI, 1.29-3.19), and light smokers classified as smoking <20 cigarettes/day, also demonstrated a higher risk of ONFH when compared with nonsmokers (OR 1.73; 95% CI, 1.06-2.83). When smoking was classified by pack-years, heavy smokers (>20 pack-years) were at a higher risk of ONFH (OR 2.26; 95% CI, 1.24-4.13), but no significant difference in risk was identified in light smokers (<20 pack-years) (OR 1.81; 95% CI, 0.88-3.71) when compared with nonsmokers. CONCLUSIONS: Our meta-analysis showed that current smokers were at a higher risk of ONFH, this high risk can also be found in former smokers. And heavy cigarette smoking showed a higher risk of ONFH than light smoking.

Detection of melamine adulteration in milk by near-infrared spectroscopy and one-class partial least squares.

Melamine is a noxious nitrogen-rich substance and has been illegally adulterated in milk to boost the protein content. The present work investigated the feasibility of using near-infrared (NIR) spectrum and one-class partial least squares (OCPLS) for detecting the adulteration of melamine. A total of 102 liquor milks were prepared for experiment. A special variable importance (VI) index was defined to select 40 most significant variables. Thirty-two pure milk samples constitute the training set for constructing a one-class model and the other samples were used for the test set. The results showed that on the independent test set, it can achieve an acceptable performance, i.e., the total accuracy of 89%, the sensitivity of 90%, and the specificity of 88%. It seems that the combination of NIR spectroscopy and OCPLS classifier can serve as a potential tool for rapid and on-site screening melamine in milk samples.

Continuous wavelet transform-based feature selection applied to near-infrared spectral diagnosis of cancer.

Spectrum is inherently local in nature since it can be thought of as a signal being composed of various frequency components. Wavelet transform (WT) is a powerful tool that partitions a signal into components with different frequency. The property of multi-resolution enables WT a very effective and natural tool for analyzing spectrum-like signal. In this study, a continuous wavelet transform (CWT)-based variable selection procedure was proposed to search for a set of informative wavelet coefficients for constructing a near-infrared (NIR) spectral diagnosis model of cancer. The CWT provided a fine multi-resolution feature space for selecting best predictors. A measure of discriminating power (DP) was defined to evaluate the coefficients. Partial least squares-discriminant analysis (PLS-DA) was used as the classification algorithm. A NIR spectral dataset associated to cancer diagnosis was used for experiment. The optimal results obtained correspond to the wavelet of db2. It revealed that on condition of having better performance on the training set, the optimal PLS-DA model using only 40 wavelet coefficients in 10 scales achieved the same performance as the one using all the variables in the original space on the test set: an overall accuracy of 93.8%, sensitivity of 92.5% and specificity of 96.3%. It confirms that the CWT-based feature selection coupled with PLS-DA is feasible and effective for constructing models of diagnostic cancer by NIR spectroscopy.

Near-infrared spectroscopy as a diagnostic tool for distinguishing between normal and malignant colorectal tissues.

Cancer diagnosis is one of the most important tasks of biomedical research and has become the main objective of medical investigations. The present paper proposed an analytical strategy for distinguishing between normal and malignant colorectal tissues by combining the use of near-infrared (NIR) spectroscopy with chemometrics. The successive projection algorithm-linear discriminant analysis (SPA-LDA) was used to seek a reduced subset of variables/wavenumbers and build a diagnostic model of LDA. For comparison, the partial least squares-discriminant analysis (PLS-DA) based on full-spectrum classification was also used as the reference. Principal component analysis (PCA) was used for a preliminary analysis. A total of 186 spectra from 20 patients with partial colorectal resection were collected and divided into three subsets for training, optimizing, and testing the model. The results showed that, compared to PLS-DA, SPA-LDA provided more parsimonious model using only three wavenumbers/variables (4065, 4173, and 5758 cm(-1)) to achieve the sensitivity of 84.6%, 92.3%, and 92.3% for the training, validation, and test sets, respectively, and the specificity of 100% for each subset. It indicated that the combination of NIR spectroscopy and SPA-LDA algorithm can serve as a potential tool for distinguishing between normal and malignant colorectal tissues.

Diagnosis of colorectal cancer by near-infrared optical fiber spectroscopy and random forest.

Near-infrared (NIR) spectroscopy has such advantages as being noninvasive, fast, relatively inexpensive, and no risk of ionizing radiation. Differences in the NIR signals can reflect many physiological changes, which are in turn associated with such factors as vascularization, cellularity, oxygen consumption, or remodeling. NIR spectral differences between colorectal cancer and healthy tissues were investigated. A Fourier transform NIR spectroscopy instrument equipped with a fiber-optic probe was used to mimic in situ clinical measurements. A total of 186 spectra were collected and then underwent the preprocessing of standard normalize variate (SNV) for removing unwanted background variances. All the specimen and spots used for spectral collection were confirmed staining and examination by an experienced pathologist so as to ensure the representative of the pathology. Principal component analysis (PCA) was used to uncover the possible clustering. Several methods including random forest (RF), partial least squares-discriminant analysis (PLSDA), K-nearest neighbor and classification and regression tree (CART) were used to extract spectral features and to construct the diagnostic models. By comparison, it reveals that, even if no obvious difference of misclassified ratio (MCR) was observed between these models, RF is preferable since it is quicker, more convenient and insensitive to over-fitting. The results indicate that NIR spectroscopy coupled with RF model can serve as a potential tool for discriminating the colorectal cancer tissues from normal ones.

The diagnostics of diabetes mellitus based on ensemble modeling and hair/urine element level analysis.

The aim of the present work focuses on exploring the feasibility of analyzing the relationship between diabetes mellitus and several element levels in hair/urine specimens by chemometrics. A dataset involving 211 specimens and eight element concentrations was used. The control group was divided into three age subsets in order to analyze the influence of age. It was found that the most obvious difference was the effect of age on the level of zinc and iron. The decline of iron concentration with age in hair was exactly consistent with the opposite trend in urine. Principal component analysis (PCA) was used as a tool for a preliminary evaluation of the data. Both ensemble and single support vector machine (SVM) algorithms were used as the classification tools. On average, the accuracy, sensitivity and specificity of ensemble SVM models were 99%, 100%, 99% and 97%, 89%, 99% for hair and urine samples, respectively. The findings indicate that hair samples are superior to urine samples. Even so, it can provide more valuable information for prevention, diagnostics, treatment and research of diabetes by simultaneously analyzing the hair and urine samples.