Search

Breast density may be used as a predictor of breast cancer risk and can measure the condition of tissues on mammograms. This research developed a computer-aided diagnosis (CAD) system to predict br...

Emergency braking intention detection has a vital practical value for improving driving safety. This paper proposed an electromyography (EMG)-based method to detect emergency braking intention from soft braking and normal driving intentions. Temporal and spectral signatures of EMG signals of emergency braking, soft braking, and normal driving intentions were investigated. Common spatial pattern (CSP) was used to generate virtual channels. The power spectrum density and linear envelope amplitude of EMG signals were used as features, respectively. Chi-square test (Chi) was used to select features. Regularized linear discrimination analysis was developed to detect emergency braking intention from the other two driving intentions. Experiment results showed significant differences in temporal and spectral domains between three kinds of driving. Furthermore, on average, the proposed method based on spectral features can detect emergency braking intention 155.70 ms before behavior under emergency situations with a system accuracy of 95.72%. The proposed method based on EMG signals for predicting emergency braking intention can be applied to develop active driving assistance systems and improve driving safety.

The sign language is developed with an intention to communicate with mute people. In the case of English or Tamil, there is a common grammatical pattern for the usage of the language. But in the case of the sign language, there is no specific grammatical pattern. Different hand gestures and body languages are used to convey a single concept in different places. Meanwhile, in order to communicate, they use the linguistic properties of a language. There is a misconception that all sign languages are same throughout the world. But, in fact, they are unique. The sign language has variations in different regions on the basis of its vocabulary and grammar. The sign language differs in different countries and even at certain places within a country. Such a phenomenon can be experienced in the case of dialectical variations of spoken languages. The mute community of southern region of India, especially Tamil Nadu, uses Tamil Sign Language (TSL) and English Sign Language (ESL) for communication purposes. The primary focus of this research is to develop a recognition system for deaf and dumb people using image processing technique. However, it seems that, the proposed gesture recognizer is a complete sign language recognizer.

It is hard to remove sweet clover seeds from alfalfa seed lots by conventional methods, affecting the purity of seed lots and resulting losses in for alfalfa hay production as well as seed yield. However, the discrimination of sweet clover seed contaminates in alfalfa seed lots is difficult without special training. In this study, multispectral imaging with object-wise multivariate image analysis was evaluated for its potential to separate sweet clover and alfalfa seeds. Principal component analysis (PCA), linear discrimination analysis (LDA), partial least squares discriminant analysis (PLSDA), AdaBoost and support vector machine (SVM) methods were applied to classify seeds of sweet clover and alfalfa according to their morphological features and spectral traits or a combination thereof. The results showed that an excellent classification could be achieved based on a combination of morphological features and spectral data in a tested data set. Seed classification accuracy was up to 99.58% in a validation set with the LDA model, which was better than the PLSDA (68.19%), AdaBoost (96.95%) and SVM (98.47%) models. Thus, multispectral imaging together with chemometric multivariate analysis is a promising technique to identify sweet clover seeds in alfalfa seed lots with high efficiency.

Great attention has been directed towards developing rapid and straightforward methods for the identification of various pesticides that are usually used simultaneously in citrus fruits. The extensive use of diverse classes of pesticides in citrus fruits and their high toxicity may cause serious diseases in the human body. In the current study, a non-enzymatic sensor array has been developed for the identification and discrimination of five different pesticides belonging to diverse classes, including organophosphate, carbamate, and bipyridylium. For this aim, two gold nanoparticles (AuNPs) with different capping agents, citrate and borohydride, were used as sensing elements. The aggregation-induced spectra alterations of AuNPs were utilized to identify the pesticides in a wide range of concentrations (20-5000 ng mL-1). We have employed data visualization methods (i.e., heat maps, bar plots, and color difference maps), a supervised pattern recognition method (i.e., linear discrimination analysis), and partial least squares regression to qualitatively and quantitatively determine the pesticides. Finally, the practical applicability of the developed sensor array was evaluated for the identification of target pesticides in lime peel. The outcomes revealed that the probe could accurately verify the absence or presence of the pesticides in lime fruit.

Obsessive-compulsive disorder (OCD) is a mental illness causing patients to suffer from recurring undesirable thoughts (obsessions) conducting to do affairs repetitively (compulsions). Brain signals recorded by Electroencephalogram (EEG) can be analyzed in order to present a diagnostic procedure considering the localization approach. In this study, the signals acquired by EEG have been recorded from three groups; two case groups; patients with severe obsessive symptoms and patients with severe compulsive symptoms, and one healthy control group. Brain signal processing techniques have been applied on the signals emitted from frontal and parieto-occipital regions to discover the features leading to the best discrimination between case groups and healthy controls. In this regard, after preprocessing, the features of time and frequency domains presenting the significant meaningful relation were nominated for classification by linear discrimination analysis (LDA). Although the parieto-occipital region performed better in the diagnosis for both obsessive and compulsive groups, the features gained from the frontal cortex resulted in better discrimination for only the compulsive group. In addition, time domain features had a more significant influence in diagnosis rather than frequency domain for both case groups. The study presented particular characteristics of brain signals in two dimensions of OCD in specific brain regions leading to more accurate presurgical assessments in the studies between the affected brain regions and behavioral issues.

Seed aging detection and viable seed prediction are of great significance in alfalfa seed production, but traditional methods are disposable and destructive. Therefore, the establishment of a rapid and non-destructive seed screening method is necessary in seed industry and research. In this study, we used multispectral imaging technology to collect morphological features and spectral traits of aging alfalfa seeds with different storage years. Then, we employed five multivariate analysis methods, i.e., principal component analysis (PCA), linear discrimination analysis (LDA), support vector machines (SVM), random forest (RF) and normalized canonical discriminant analysis (nCDA) to predict aged and viable seeds. The results revealed that the mean light reflectance was significantly different at 450~690 nm between non-aged and aged seeds. LDA model held high accuracy (99.8~100.0%) in distinguishing aged seeds from non-aged seeds, higher than those of SVM (87.4~99.3%) and RF (84.6~99.3%). Furthermore, dead seeds could be distinguished from the aged seeds, with accuracies of 69.7%, 72.0% and 97.6% in RF, SVM and LDA, respectively. The accuracy of nCDA in predicting the germination of aged seeds ranged from 75.0% to 100.0%. In summary, we described a nondestructive, rapid and high-throughput approach to screen aged seeds with various viabilities in alfalfa.

The illegal disposal of industrial sewage sludge has caused serious environmental pollution. To develop identification technology of industrial sewage sludge based on the characteristic fingerprints is a promising method that is helpful to clarify the responsibility of illegal enterprises. In this study, heavy metal profiles of sewage sludge from industries (including printing and dyeing industry and other industries) and municipal sewage treatment plant located in eastern China were determined, and their performance of classification was evaluated by principal component analysis (PCA) and linear discrimination analysis (LDA). Results showed that heavy metal composition can be an effective tool for distinguishing sewage sludge between printing and dyeing industry and other industries, with an accuracy rate of 82.9%. Meanwhile, heavy metal speciation may be a promising method for identification of printing and dyeing sludge from municipal sewage sludge, the accuracy rate of which reached 100%. Moreover, antimony (Sb) and zinc (Zn) are two indicators, which can be used to identify sewage sludge between printing and dyeing sub-industries, and the accuracy rate was 90%. We concluded that heavy metal profiles may be a precise and promising tool for identification of printing and dyeing sludge. This study developed a potential method for tracing the source of industrial sewage sludge and establishing the identification database of industrial sewage sludge and provided technical support for the government to supervise the illegal dumping and disposal of industrial sewage sludge.

The research work proposes a system which can handle different types of selected words into a common vision-based platform. These sign languages are used not only to communicate with deaf and dumb people but also to exchange ideas with people who can hear but physically cannot speak. Most of the researchers work on the sign language by processing the video with frames in an equal or regular interval. Some of the factors that affect this phenomenon are colour, angle, the position of hand-disturb, dialectical variations of languages etc. All these factors act as impediments to create a perfect sign language system. Region based analysis is used to exploit the interior pixels and boundary of the object. These shape descriptors are more sensible to distortions and noise. The sign language learning is classified into unsupervised and supervised learning. The studies in the realm of sign language can be comprehended as a complicated one and researches are going on in the fields of supervised and unsupervised learning. The research work proposes a system which can handle different types of selected words into a common vision-based platform. Feature extractions like Wavelet transform method can be used in future for better results. Other classifiers like Component Analysis, Support Vector Machine and Linear Discrimination Analysis can also be used for conducting experiments and to improve the recognition rate.

Two-dimensional transition metal carbides/nitrides, known as MXenes, have recently received significant attention for gas sensing applications. However, Mxenes have strong adsorption to many types of volatile organic compounds (VOCs), and therefore gas sensors based on MXenes generally have poor selectivity. Herein, we developed a Ti 3 C 2 T X based multivariable sensor which allows identification of different VOCs, and concentration prediction of the target VOC in variable backgrounds. The sensor’s responses from the broadband impedance spectra create a unique fingerprint for each VOC. Based on the methodologies of principal component analysis and linear discrimination analysis, we demonstrate highly accurate identifications for different types of VOCs and mixtures using this multivariable sensor. Furthermore, we demonstrate an accuracy of 93.2% for the prediction of ethanol concentrations in the presence of different concentrations of water and methanol. The high level of identification and concentration estimation shows the great potential of MXene based multivariable sensor for detection of target VOCs in the presence of known and unknown interferences.

In the market, two types of honey are available: monofloral and multifloral honeys. Monofloral honey has been regarded having higher market values than multifloral one. In order to protect a fraud trading between monofloral and multifloral honeys, it is very important to develop an analytical method which can be used to discriminate the two type of honeys. We utilized spectral data in the UV-visible region (250-450 nm) coupled with linear discrimination analysis (LDA) to classify monofloral and multifloral honeys from Indonesia. Total 400 samples of monofloral and multifloral honeys were used as samples. The spectral data were recorded using UV-vis spectrometer in the wavelength of 190-1100 nm with 1 nm of resolution. Several preprocessing methods was applied to improve the quality of spectral data. Principal component analysis (PCA) was calculated to map the samples and the PCA scores were used as input for classification task using linear discriminant analysis (LDA). The result suggested that UV-Visible spectroscopy is a powerful tool for quality evaluation of Indonesian honeys.

Simple, rapid, and accurate detection methods for saccharides are potentially applicable to various fields such as clinical and food chemistry. However, the practical applications of on-site analytical methods are still limited. To this end, herein, we propose a 96-well microtiter plate made of paper as a paper-based chemosensor array device (PCSAD) for the simultaneous classification of 12 saccharides and the quantification of fructose and glucose among 12 saccharides. The mechanism of the saccharide detection relied on an indicator displacement assay (IDA) on the PCSAD using four types of catechol dyes, 3-nitrophenylboronic acid, and the saccharides. The design of the PCSAD and the experimental conditions for the IDA were optimized using a central composite design. The chemosensors exhibited clear color changes upon the addition of saccharides on the paper because of the competitive boronate esterification. The color changes were employed for the subsequent qualitative, semiquantitative, and quantitative analyses using an automated algorithm combined with pattern recognition for digital images. A qualitative linear discrimination analysis offered discrimination of 12 saccharides with a 100% classification rate. The semiquantitative analysis of fructose in the presence of glucose was carried out from the viewpoint of food analysis utilizing a support vector machine, resulting in clear discrimination of the various concentrations of fructose. Most importantly, the quantitative detection of fructose in two types of commercial soft drinks was also successfully carried out without sample pretreatments. Thus, the proposed PCSAD can be a powerful method for on-site food analyses that can meet the increasing demand from consumers for sensors of saccharides.

We propose a colorimetric sensing array consisting of 4-aminothiophenol (p-ATP)-modified gold nanoparticles (Au NPs), silver nanoparticles (Ag NPs), and core-shell Au@Ag nanocubes (Au@Ag NCs) as sensing elements to identify multiple proteins according to the diverse colorimetric response patterns. In the absence of proteins, the sensor element solution itself did not agglomerate. After interacting with six proteins (lysozyme (LZM), hemoglobin (HGB), peroxidase from horseradish (HRP), bovine liver from peroxidase (CAT), trypsin from bovin pancreas (TRY), and pepsin (PEP)), due to the different binding ability between the sensing elements and various proteins, the sensing array exhibits a unique pattern of colorimetric variations, linear discrimination analysis (LDA) was applied to analyze the pattern and produced a clustering map for a clearer differentiation of these proteins.

To rapidly and accurately distinguish the real and fake bloods, a photoacoustic detection system was established based on an optical parameters oscillator (OPO) pulsed laser and a focused ultrasonic transducer in this paper. The time-resolved photoacoustic signals of ten cow bloods and ten props bloods were captured. The amplitudes and profiles of photoacoustic signals for real and fake bloods were different. In addition, the photoacoustic peak-to-peak values of all blood samples were obtained from 700nm to 1064nm, and used as the train samples. The photoacoustic peak-to-peak values were firstly standardized and the data dimensions were reduced by using principle component analysis (PCA) method. Different principle components were chosen according to the cumulative contribution rate of characteristic values for the standardized photoacoustic peak-to-peak data matrix. Then, the scores of different principle components for ten cow bloods and ten props bloods were obtained and used as the train data. Five cow bloods and five props bloods were test. The linear discrimination analysis (LDA) algorithm was used to distinguish the real and fake bloods. The test results show that, based on the score data for two principle components by using PCA, the correct rate (CR) of real and fake bloods was improved from 80% to 100%. Results show that the LDA coupled with PCA, i.e.,PCA-LDA, has good distinction effect for the real and fake bloods.

Backgrounds and Aims There is no accurate and reliable circulating biomarker to diagnose Crohn’s disease [CD]. Raman spectroscopy is a relatively new approach that provides information on the biochemical composition of samples in minutes and virtually without any sample preparation. We aimed to test the use of Raman spectroscopy analysis of plasma samples as a potential diagnostic tool for CD. Methods We analysed by Raman spectroscopy dry plasma samples obtained from 77 CD patients [CD] and 45 healthy controls [HC]. In the dataset obtained, we analysed spectra differences between CD and HC, as well as among CD patients with different disease behaviours. We also developed a method, based on principal component analysis followed by a linear discrimination analysis [PCA-LDA], for the automatic classification of individuals based on plasma spectra analysis. Results Compared with HC, the CD spectra were characterised by less intense peaks corresponding to carotenoids [p Conclusions Our results indicate that Raman spectroscopy of blood plasma can identify metabolic variations associated with CD and it could be a rapid pre-screening tool to use before further specific evaluation.

Traceability of honey is highly required by consumers and food administration with the consideration of food safety and quality. In this study, a technique named laser-induced breakdown spectroscopy (LIBS) was used to fast trace geographical origins of acacia honey and multi-floral honey. LIBS emissions from elements of Mg, Ca, Na, and K had significant differences among different geographical origins. The clusters of honey from different geographical origins were visualized with principal component analysis. In addition, support vector machine (SVM) and linear discrimination analysis (LDA) were used to quantitively classify the origins. The results indicated that SVM performed better than LDA, and the discriminant results of multi-floral honey were better than acacia honey. The accuracy and mean average precision for multi-floral honey were 99.7% and 99.7%, respectively. This study provided a fast approach for geographical origin classification, and might be helpful for food traceability.

The number of different types of cheese worldwide exceeds 4000 and dairy fat, composed of about 400 fatty acids (FA), is one of the most complex dietary fats. Cheeses are valuable sources of different bioactive FA, i.e., conjugated FA (CFA). The aim of present study was to determine FA profile of commercially available ripening cheeses, with the special emphasis on CFA profile. Multivariate analyses (cluster analysis (CA), principal component Analysis (PCA), and linear discriminant analysis (LDA)) of chromatographic data have been proposed as an objective approach for evaluation and data interpretation. CA enabled the differentiation of ripening cheeses from fresh cheeses and processed cheeses. PCA allowed to differentiate some types of ripening cheese whereas proposed LDA model, based on 22 analyzed FA, enabled assessing cheeses type with average predictive sensitivities of 86.5%. Results of present study clearly demonstrated that FA and CFA content may not only contribute to overall nutritional characteristics of cheese but also, when coupled with chemometric techniques, may be used as chemical biomarkers for assessing the origin and/or the type of ripening cheeses and the confirmation of their authenticity, which is of utmost importance for consumers.

In this paper, to address driving safety under emergency situations, we investigated the neural signatures of emergency braking and soft braking of drivers and proposed recognition methods to distinguish the two kinds of braking intentions by using electroencephalography (EEG) signals. The neural signatures are captured in the temporal and spectral domains of EEG signals. Temporal features, spectral features, and combined features were used as input features of the recognition model based on the regularized linear discrimination analysis. Experimental results showed significant differences in neural signatures between emergency braking and soft braking, and the recognition model can detect the emergency braking and soft braking intentions from the normal driving intention well. These features were extracted from response-locked segments. The recognition accuracy for the three-class classification reached 78.33% ± 6.86% by using the system based on combined features. This works can further advance the development of EEG-based human-centric assistant driving systems to improve driving safety and driving comfort.

The objectives of this study are to identify the significant variables and to verify the best statistical method for determining the effect of indoor air quality (IAQ) at 7 different locations in Universiti Sultan Zainal Abidin, Terengganu, Malaysia. The IAQ data were collected using in-situ measurement. Principal component analysis (PCA), partial least squares discriminant analysis (PLS-DA), linear discrimination analysis (LDA), and agglomerative hierarchical clustering (AHC) were used to classify the significant variables as well as to compare the best method for determining IAQ levels. PCA verifies only 4 out of 9 parameters (PM10, PM2.5, PM1.0, and O3) and is the significant variable in IAQ. The PLS-DA model classifies 89.05% correct of the IAQ variables in each station compared to LDA with only 66.67% correct. AHC identifies three cluster groups, which are highly polluted concentration (HPC), moderately polluted concentration (MPC), and low-polluted concentration (LPC) area. PLS-DA verifies the groups produced by AHC by identifying the variables that affect the quality at each station without being affected by redundancy. In conclusion, PLS-DA is a promising procedure for differentiating the group classes and determining the correct percentage of variables for IAQ.

Reliable discrimination of the geographical origin of tea, especially very well-known teas, is crucial for market developing and consumer rights’ protection. In this study, multi-element contents and stable isotope signatures in the flat-shaped green tea samples collected from different producing areas were assayed. Linear discrimination analysis (LDA), partial least squares discrimination analysis (PLS-DA), and a decision tree (DT) were tested for their ability to discriminate the tea’s geographical origin. Under the validation by cross-validation and “blind” dataset, the prediction accuracies of the three methods were all greater than 70%. The DT method showed the best performance, with an accuracy of 90%. Furthermore, for the discrimination of Xihu Longjing (XHLJ) green tea, DT also showed the lowest error rate of 1.5% (1 of 67 wrongly classified to XHLJ), which was better than the 6% rate observed for PLS-DA.

It is evident that surface electromyography (sEMG)-based sensing approach for human-machine interfaces has some inherent limitations for applications requiring morphological changes information of deep-layered muscles, such as dexterous prosthetic hands. In this paper, the design, simulation, fabrication, and evaluation for a series of novel structured ultrasound transducers are conducted in order to develop a type of A-mode ultrasound transducers that overcome the drawbacks of the sEMG-based sensing. The transducers cover single-frequency and dual-frequency types. Their key parameters, the acoustic impedance and thickness of the matching layer, are simulated and verified by PZFlex. The parameters are designed as 0.3 times of the 1-3 composite piezoelectric's acoustic impedance and 0.25 times of the wavelength, respectively. The characterizations of the dual-frequency transducers significantly outperform single-frequency transducers. The experiments of recognizing dexterous hand gesture are designed to detect morphological changes information of deep-layered muscles. The classification accuracy improvements with linear discrimination analysis are 7.3% and 4.7%, and with support vector machine are 14.1% and 13.4% for the horizontal stacked and annulus array. This preliminary study concludes that the dual-frequency transducers have huge potential for applications that need contraction information of deep-layered muscles over the single-frequency transducers, letting alone sEMG-based sensors.

Advanced techniques were required to determine geographical origins and production methods of aquatic food due to frequent fraudulent misdescription of food content on product labels in the market. In this study, the elemental profile of 96 salmonids samples were detected by inductively coupled plasma atomic emission spectrometry (ICP-AES) and used to identify production methods and geographical origins of salmonids. Moreover, effects of fish size and harvest time on the elemental compositions in salmonids were investigated. A total of 14 elements (Al, Ca, Co, Cr, Cu, Fe, Ga, K, Mg, Mn, Na, Ni, Sr, and Zn) were determined in salmonids. Of these, elements Fe, K, Na, Mn, and Zn, were varied in fish collected in different harvest times, and elements Al, Fe, Ga, Mn, and Zn were related to fish sizes, six elements (Ga, K, Mg, Na, Ni, and Sr) were significantly different between freshwater-cultured and seawater-cultured salmonids, 13 elements (Al, Ca, Co, Cr, Fe, Ga, K, Mg, Mn, Na, Ni, Sr, and Zn) were significantly different among different areas. Three elements (K, Mg, and Na) and 12 elements (Ca, Co, Cr, Fe, Ga, K, Mg, Mn, Na, Ni, Sr, and Zn) were selected by SDA for identifying production methods and geographical origins of salmonids, respectively. Canonical discriminant analysis (CDA) visualized the difference in salmonids cultured by different production methods or from different areas. Multivariate statistics, linear discrimination analysis (LDA), k-nearest neighbor (KNN), and random forest (RF) demonstrated the validity of the elemental profile in distinguishing production methods and geographical origins of salmonids. The results showed that production methods of salmonids were distinguished by LDA, KNN, and RF with an overall accuracy of >94.74%. Among these, the highest accuracy was achieved by KNN with an accuracy of 100%. For authenticating geographical origins of salmonids, the accuracy of KNN (> 94.74%) was highest. The element profile was effective to distinguish production methods and geographical origins of salmonids, which were less affected by harvest time and fish size.