Optimal Scree-CNN for Detecting NS1 Molecular Fingerprint from Salivary SERS Spectra.

Dengue fever (DF) is a viral infection with possible fatal consequence. NS1 is a recent antigen based biomarker for dengue fever (DF), as an alternative to current serum and antibody based biomarkers. Convolutional Neural Network (CNN) has demonstrated impressive performance in machine learning problems. Our previous research has captured NS1 molecular fingerprint in saliva using Surface Enhanced Raman Spectroscopy (SERS) with great potential as an early, noninvasive detection method. SERS is an enhanced variant of Raman spectroscopy, with extremely high amplification that enables spectra of low concentration matter, such as NS1 in saliva, readable. The spectrum contains 1801 features per sample, at a total of 284 samples. Principal Component Analysis (PCA) transforms high dimensional correlated signal to a lower dimension uncorrelated principal components (PCs), at no sacrifice of the original signal content. This paper aims to unravel an optimal Scree-CNN model for classification of salivary NS1 SERS spectra. Performances of a total of 490 classifier models were examined and compared in terms of performance indicators [accuracy, sensitivity, specificity, precision, kappa] against a WHO recommended clinical standard test for DF, enzyme-linked immunosorbent assay (ELISA). Effects of CNN parameters on performances of the classifier models were also observed. Results showed that Scree-CNN classifier model with learning rate of 0.01, mini-batch size of 64 and validation frequency of 50, reported an across-the-board 100% for all performance indicators.

Reduced Featured k-NN Classifier Model Optimal for Classification of Dengue Fever from Salivary Raman Spectra.

Current diagnostic methods based on nonstructural protein 1 (NS1) for dengue infection use blood as the medium and hence are invasive. Worry for blood infected diseases, pain from pricking, overcrowded public hospitals and ignorance are just a few of the causes for delayed diagnosis that contributes to mortality from dengue fever (DF). NS1 has also been reported in saliva, but sensitivity of detection is much lower than that of blood. If saliva is to be a medium, detection of NS1 requires a more specific and sensitive technique. In this study, we are exploiting the advantages of saliva and Surface Enhanced Raman Spectroscopy (SERS) to develop a non-invasive early detection method for DF. Significant features from Raman spectra of saliva samples of dengue suspected patients and healthy volunteers were extracted with Principal Component Analysis (PCA) and served as input to k-Nearest Neighbour (k-NN) for classification. Cumulative Percentage Variance (CPV) is the criterion for feature extraction. Two k-NN distance rules (Cosine and Manhattan) combined with k-values ranging from 3 to 17 were varied to obtain an optimal k-NN classifier. Then, performance of the different k-NN classifier models is benchmarked against Panbio Dengue Early ELISA and SD BIOLINE Dengue Duo technique from the clinical laboratory. The finding is encouraging with the best performance achieved, 82.14% for accuracy, 85.71% for sensitivity and 78.57% for specificity.

Optimal ELM-RBF model and SERS Analysis of Saliva for Classification of NS1.

Extreme Learning Machine (ELM) with Radial Basis Function (RBF) Kernel has demonstrated strong capability in pattern recognition and classification problems. NS1 is a biomarker for flavivirus related diseases, where current detection methods are serum based and hence invasive. Our previous work has captured NS1 molecular fingerprint in saliva using Surface Enhanced Raman Spectroscopy (SERS) that could amount to non-invasive detection method. SERS is an improved Raman spectroscopic technique, which can amplify spectral intensity by 103 to l07 times, to yield usable spectra of low concentration NS1 in saliva. The spectra produced contain 1801 features for each of the 284 samples collected. Principal Component Analysis (PCA) transforms a high dimensional data to a lower dimension principal components (PCs), at no sacrifice of important information of the original data. Both termination criteria of PCA and kernel parameters of ELM have effect on performance of the classifier models. This paper aims to unravel an optimal ELM-RBF classifier model for classification of NS1 salivary SERS spectra. Performance of a total of 864 classifier models are examined and compared in terms of [accuracy, kappa, precision, sensitivity and specificity]. Results show that CPV- and EOC-ELM-RBF classifier models are on par and outperform the Scree-ELM-RBF classifier models.

Comparison of Extreme Learning Machine and K-Nearest Neighbour Performance in Classifying EEG Signal of Normal, Poor and Capable Dyslexic Children.

Dyslexia is a specific learning difficulty associated with brain capability in processing numbers and letters. Analysis of Electroencephalogram (EEG) could provide insight information on differences in brain processing. In this work, two machine learning techniques were applied to distinguish EEG signals of normal, poor and capable dyslexic children during writing word and non-word. The performance of k-nearest neighbour (KNN) with correlation distance function and extreme learning machine (ELM) with radial basis function (RBF) were compared. The performance of each classifier was determined using sensitivity, specificity and accuracy. It was found that ELM was capable of classifying the dyslexic children with 89% accuracy compared to KNN which is only 83%. These results showed that ELM is feasible and reliable in recognising normal, poor and capable dyslexic children through writing.

A ROIC for Mn(TPP)Cl-DOP-THF-Polyhema PVC membrane modified n-channel Si3N4 ISFET sensitive to histamine.

Intolerance of histamine could lead to scombroid poisoning with fatal consequences. Current detection methods for histamine are wet laboratory techniques which employ expensive equipment that depends on skills of seasoned technicians and produces delayed test analysis result. Previous works from our group has established that ISFETs can be adapted for detecting histamine with the use of a novel membrane. However, work to integrate ISFETs with a readout interfacing circuit (ROIC) circuit to display the histamine concentration has not been reported so far. This paper concerns the development of a ROIC specifically to integrate with a Mn(TPP)Cl-DOP-THF-Polyhema PVC membrane modified n-channel Si3N4 ISFET to display the histamine concentration. It embodies the design of constant voltage constant current (CVCC) circuit, amplification circuit and micro-controller based display circuit. A DC millivolt source is used to substitute the membrane modified ISFET as preliminary work. Input is histamine concentration corresponding to the safety level designated by the Food and Drugs Administration (FDA). Results show the CVCC circuit makes the output follows the input and keeps VDS constant. The amplification circuit amplifies the output from the CVCC circuit to the range 2.406-4.888V to integrate with the microcontroller, which is programmed to classify and display the histamine safety level and its corresponding voltage on a LCD panel. The ROIC could be used to produce direct output voltages corresponding to histamine concentrations, for in-situ applications.

PCA-MLP SVM distinction of salivary Raman spectra of dengue fever infection.

Dengue fever (DF) is a disease of major concern caused by flavivirus infection. Delayed diagnosis leads to severe stages, which could be deadly. Of recent, non-structural protein (NS1) has been acknowledged as a biomarker, alternative to immunoglobulins for early detection of dengue in blood. Further, non-invasive detection of NS1 in saliva makes the approach more appealing. However, since its concentration in saliva is less than blood, a sensitive and specific technique, Surface Enhanced Raman Spectroscopy (SERS), is employed. Our work here intends to define an optimal PCA-SVM (Principal Component Analysis-Support Vector Machine) with Multilayer Layer Perceptron (MLP) kernel model to distinct between positive and negative NS1 infected samples from salivary SERS spectra, which, to the best of our knowledge, has never been explored. Salivary samples of DF positive and negative subjects were collected, pre-processed and analyzed. PCA and SVM classifier were then used to differentiate the SERS analyzed spectra. Since performance of the model depends on the PCA criterion and MLP parameters, both are examined in tandem. Its performance is also compared to our previous works on simulated NS1 salivary samples. It is found that the best PCA-SVM (MLP) model can be defined by 95 PCs from CPV criterion with P1 and P2 values of 0.01 and -0.2 respectively. A classification performance of [76.88%, 85.92%, 67.83%] is achieved.

PCA criterion for SVM (MLP) classifier for flavivirus biomarker from salivary SERS spectra at febrile stage.

Non-structural protein (NS1) has been conceded as one of the biomarkers for flavivirus that causes diseases with life threatening consequences. NS1 is an antigen that allows detection of the illness at febrile stage, mostly from blood samples currently. Our work here intends to define an optimum model for PCA-SVM with MLP kernel for classification of flavivirus biomarker, NS1 molecule, from SERS spectra of saliva, which to the best of our knowledge has never been explored. Since performance of the model depends on the PCA criterion and MLP parameters, both are examined in tandem. Input vector to classifier determined by each PCA criterion is subjected to brute force tuning of MLP parameters for entirety. Its performance is also compared to our previous works where a Linear and RBF kernel are used. It is found that the best PCA-SVM (MLP) model can be defined by 5 PCs from Cattel's Scree test for PCA, together with P1 and P2 values of 0.1 and -0.2 respectively, with a classification performance of [96.9%, 93.8%, 100.0%].

Model Selection for PCA-Linear SVM for automated detection of NS1 molecule from Raman spectra of salivary mixture.

Of recent, detection of Non-structural Protein 1 (NS1) in saliva has become appealing, as it may lead to a noninvasive detection method for NS1-related diseases at the febrile phase, before complication developed. NS1 is found to have a molecular fingerprint with the use of SERS technique. Our work here intends to determine an optimum PCA-Linear SVM model for automated detection of NS1 molecules from Raman spectra of NS1 adulterated saliva. Raman spectra of normal saliva (n=64) and saliva adulterated with low concentration NS1 (n=64) are used. Since Raman features extracted for each spectrum numbered at 1801, ranking and selection of features in order of their contribution is important prior to classification, for efficient computation. Hence, PCA for feature selection and SVM with linear kernel for classification are integrated. It is found that the Cattel's Scree test is the best stopping criteria for PCA with a selection of 5 PCs and a box constraint of 20 is optimum for Linear SVM. Together they achieve a classification performance, [accuracy sensitivity, specificity], of [98.71% 98.97% 98.44%].

EEG-based time and spatial interpretation of activation areas for relaxation and words writing between poor and capable dyslexic children.

Symptoms of dyslexia such as difficulties with accurate and/or fluent word recognition, and/or poor spelling as well as decoding abilities, are easily misinterpreted as laziness and defiance amongst school children. Indeed, 37.9% of 699 school dropouts and failures are diagnosed as dyslexic. Currently, Screening for dyslexia relies heavily on therapists, whom are few and subjective, yet objective methods are still unavailable. EEG has long been a popular method to study the cognitive processes in human such as language processing and motor activity. However, its interpretation is limited to time and frequency domain, without visual information, which is still useful. Here, our research intends to illustrate an EEG-based time and spatial interpretation of activated brain areas for the poor and capable dyslexic during the state of relaxation and words writing, being the first attempt ever reported. From the 2D distribution of EEG spectral at the activation areas and its progress with time, it is observed that capable dyslexics are able to relax compared to poor dyslexics. During the state of words writing, neural activities are found higher on the right hemisphere than the left hemisphere of the capable dyslexics, which suggests a neurobiological compensation pathway in the right hemisphere, during reading and writing, which is not observed in the poor dyslexics.

Classification of salivary based NS1 from Raman Spectroscopy with support vector machine.

Non-Structural Protein 1 (NS1) antigen has been recognized as a biomarker for diagnosis of flavivirus viral infections at early stage. Surface Enhanced Raman Spectroscopy (SERS) is an optical technique capable of detecting up to a single molecule. Our previous work has established the Raman fingerprint of NS1 with gold as substrate. Our current study aims to classify NS1 infected saliva samples from healthy samples, a first ever attempt. Saliva samples from healthy subjects, NS1 protein and NS1-saliva mixture samples were analyzed using SERS. The SERS spectra were then pre-processed prior to classification with support vector machine (SVM). NS1-saliva mixture at concentration of 10ppm, 50ppm and 100ppm were examined. Performance of SVM classifier with linear, polynomial and radial basis function (RBF) kernels were compared, in term of accuracy, sensitivity, and specificity. From the results, it can be concluded that SVM classifier is able to classify the samples into NS1 infected samples and normal saliva samples. Of the three kernels, performance in using polynomial and RBF kernel is found surpassing the linear kernel. The best performance is attained with RBF kernel with accuracy of [97.1% 93.4% 81.5%] for 100ppm, 50ppm and 10ppm respectively.

Raman molecular fingerprint of non-structural protein 1 in phosphate buffer saline with gold substrate.

SERS is a form of Raman spectroscopy that is enhanced with nano-sensing chip as substrate. It can yield distinct biochemical fingerprint for molecule of solids, liquids and gases. Vice versa, it can be used to identify unknown molecule. It has further advantage of being non-invasive, non-contact and cheap, as compared to other existing laboratory based techniques. NS1 has been clinically accepted as an alternative biomarker to IgM in diagnosing viral diseases carried by virus of flaviviridae. Its presence in the blood serum at febrile stage of the flavivirus infection has been proven. Being an antigen, it allows early detection that can help to reduce the mortality rate. This paper proposes SERS as a technique for detection of NS1 from its scattering spectrum. Contribution from our work so far has never been reported. From our experiments, it is found that NS1 protein is Raman active. Its spectrum exhibits five prominent peaks at Raman shift of 548, 1012, 1180, 1540 and 1650 cm(-1). Of these, peak at 1012 cm(-1) scales the highest intensity. It is singled out as the peak to fingerprint the NS1 protein. This is because its presence is verified by the ring breathing vibration of the benzene ring structure side chain molecule. The characteristic peak is found to vary in proportion to concentration. It is found that for a 99% change in concentration, a 96.7% change in intensity is incurred. This yields a high sensitivity of about one a.u. per ppm. Further investigation from the characterization graph shows a correlation coefficient of 0.9978 and a standard error estimation of 0.02782, which strongly suggests a linear relationship between the concentration and characteristic peak intensity of NS1. Our finding produces favorable evidence to the use of SERS technique for detection of NS1 protein for early detection of flavivirus infected diseases with gold substrate.

Nonstructural protein 1 characteristic peak from NS1-saliva mixture with Surface-Enhanced Raman spectroscopy.

Surface Enhanced Raman spectroscopy (SERS) is an enhanced technique of Raman spectroscopy, which amplifies the intensity of Raman scattering to a practical range with adsorption of analyte onto nano-size plasmonic material such as gold, silver or copper. This feature of SERS has given it a niche in tracing molecular structure, especially useful for marking diseases specific biomarker. NS1 protein has been clinically accepted as an alternative biomarker for diseases caused by flavivirus. Detection of Nonstructural Protein 1 (NS1) will allow early diagnosis of the diseases. Its presence in the blood serum has been reported as early as first day of infection. With gold substrate, our work here intends to explore if SERS is suitable to detect NS1 from saliva, with saliva becoming the most favored alternative to blood as diagnostic fluid due to its advantages in sample collection. Our experimental results find both gold coated slide (GS) and saliva being Raman inactive, but the molecular fingerprint of NS1 protein at Raman shift 1012 cm(-1), which has never been reported before. The distinct peak is discovered to be attributed by breathing vibration of the benzene ring structure of NS1 side chain molecule. The characteristic peak is also found to vary in direct proportion to concentration of the NS1-saliva mixture, with a correlation coefficient of +0.96118 and a standard error estimation of 0.11382.

Wavelet packet analysis of EEG signals from dyslexic children with writing disability.

This paper describes Wavelet Packet Analysis of EEG signal of dyslexic children with writing disability. Two activities were carried out during EEG recordings; relax and and writing letters. EEG signals were collected using biosignal gMobilab system and analysed using Wavelet Packet Decomposition to extract alpha and beta brainwave rhythm. Statistical data such as log energy entropy and standard deviation were used to compare the characteristic of EEG signals from dyslexic and normal children. Result showed that the dyslexic children consumed higher energy at left parietal lobe during writing activity especially those who write incorrectly. The alpha band shows higher log energy entropy for dyslexic children compare to normal children at most channel during relax.