Non-invasive Characterization of Glycosuria and Identification of Biomarkers in Diabetic Urine Using Fluorescence Spectroscopy and Machine Learning Algorithm.

The current study presents a steadfast, simple, and efficient approach for the non-invasive determination of glycosuria of diabetes mellitus using fluorescence spectroscopy. A Xenon arc lamp emitting light in the range of 200-950 nm was used as an excitation source for recording the fluorescent spectra from the urine samples. A consistent fluorescence emission peak of glucose at 450 nm was found in all samples for an excitation wavelength of 370 nm. For confirmation and comparison, the fluorescence spectra of non-diabetic (healthy controls) were also acquired in the same spectral range. It was found that fluorescence emission intensity at 450 nm increases with increasing glucose concentration in urine. In addition, optimized synchronous fluorescence emission at 357 nm was used for simultaneously determining a potential diabetes biomarker, Tryptophan (Trp) in urine. It was also found that the level of tryptophan decreases with the increase in urinary glucose concentration. The quantitative estimation of urinary glucose can be demonstrated based on the intensity of emission light carried by fluorescence light. Moreover, the dissimilarities were further emphasized using the hierarchical cluster analysis (HCA) algorithm. HCA gives an obvious separation in terms of dendrogram between the two data sets based on characteristic peaks acquired from their fluorescence emission signatures. These results recommend that urinary glucose and tryptophan fluorescence emission can be used as potential biomarkers for the non-invasive analysis of diabetes.

Thermal Effects on Biochemical Signatures of UHT, Pasteurized and Domestically Boiled Buffalo Milk Detected by Synchronous Fluorescence Spectroscopy.

Thermal treatment of milk is performed to limit bacterial growth and make it safe for human consumption. To increase the shelf life of milk, either ultrahigh temperature (UHT) or pasteurization techniques are employed that destroy the microorganisms. In this study, the synchronous front face fluorescence spectroscopy was employed for comparative study of raw, UHT treated, pasteurized and conventionally boiled milk at 93 °C (domestic boiling). Principal Component analysis clearly showed distinct clustering of UHT milk due to formation of Maillard reaction products. Fluorescence emission peak at 410 nm showed irreversible change in peak intensity attributed to conformational changes in protein due to UHT treatment while pasteurization and domestic boiling showed reversible changes when milk was cooled down to 10 °C. Furthermore, fluorescence emission peaks at 410 nm previously assigned to vitamin A has also been discussed.

Utilizing machine learning algorithms for precise discrimination of glycosuria in fluorescence spectroscopic data.

Fluorescence spectroscopy coupled with a random forest machine learning algorithm offers a promising non-invasive approach for diagnosing glycosuria, a condition characterized by excess sugar in the urine of diabetic patients. This study investigated the ability of this method to differentiate between diabetic and healthy control urine samples. Fluorescent spectra were captured from urine samples using a Xenon arc lamp emitting light within the 200 to 950 nm wavelength range, with consistent fluorescence emission observed at 450 nm under an excitation wavelength of 370 nm. Healthy control samples were also analyzed within the same spectral range for comparison. To distinguish spectral differences between healthy and infected samples, the random forest (RF) and K-Nearest Neighbors (KNN) machine learning algorithms have been employed. These algorithms automatically recognize spectral patterns associated with diabetes, enabling the prediction of unknown classifications based on established samples. Principal component analysis (PCA) was utilized for dimensionality reduction before feeding the data to RF and KNN for classification. The model's classification performance was evaluated using 10-fold cross-validation, resulting in the proposed RF-based model achieving accuracy of 96 %, specificity of 100 %, sensitivity of 93 %, and precision of 100 %. These results suggest that the proposed method holds promise for a more convenient and potentially more accurate method for diagnosing glycosuria in diabetic patients.

Integrating Laser-Induced Breakdown Spectroscopy and Ensemble Learning as Minimally Invasive Optical Screening for Diabetes.

Diabetes mellitus is a prevalent chronic disease necessitating timely identification for effective management. This paper introduces a reliable, straightforward, and efficient method for the minimally invasive identification of diabetes mellitus through nanosecond pulsed laser-induced breakdown spectroscopy (LIBS) by integrating a state-of-the-art machine learning approach. LIBS spectra were collected from urine samples of diabetic and healthy individuals. Principal component analysis and an ensemble learning classification model were used to identify significant changes in LIBS peak intensity between the diseased and normal urine samples. The model, integrating six distinct classifiers and cross-validation techniques, exhibited high accuracy (96.5%) in predicting diabetes mellitus. Our findings emphasize the potential of LIBS for diabetes mellitus identification in urine samples. This technique may hold potential for future applications in diagnosing other health conditions.

Evaluating the performance of multilayer perceptron algorithm for tuberculosis disease Raman data.

In spite of developments in various molecular approaches, major challenges remain in rapidly diagnosing infectious diseases triggered by bacteria. Identification of such causative pathogens at an earlier stage and with an acceptable degree of sensitivity and specificity would play a major role in initiating proper treatment. In this study the performance of multilayer perceptron (MLP) algorithm on the Raman Spectroscopic data of tuberculosis disease have been evaluated. Blood sera samples of TB positive (active patients), TB negative (recovered) and control (healthy) are analyzed in current study. Classifications among the data sets are based on the differences/similarities in Raman peak intensity. The analysis has been carried out by using MLP, a class of artificial neural network algorithm. The results of these classifications are built on intensities of most dominated Raman peaks i.e. 1001, 1152, 1282, 1430, 1475, and 1690cm-1. These Raman shifts are attributed to biomolecules concentration such as phenylalanine, ß-carotene, amide III and C=O of amide-I band of protein etc. The performance of the proposed model is evaluated using 5-fold cross validation method for the data sets i.e. control vs. TB positive, control vs. TB negative and TB positive vs. TB negative. The sensitivity and specificity predicted by the model is in the range of 62-92% and 81-88%, respectively. Once trained on known data set, Raman spectroscopy together with statistical algorithms can provide real time prediction for unknown samples.

Contrast enhancement in widefield CARS microscopy by tailored phase matching using a spatial light modulator.

We introduce a widefield CARS microscope implementation that uses a spatial light modulator to obtain extremely precise control over the pump/probe-beam incidence geometry, which provides the possibility to enhance the image contrast at specific target resonances by fine-tuning the incidence angles. We show how this technique can be used to optimize the image contrast between objects of different size and to practically eliminate the undesired signal from the solvent that embeds small target specimens. Changing the numerical aperture of the illumination from 1.27 to 1.24 improved the ratio of the signals of 500 nm polystyrene beads and the agarose solvent by about 20 dB.

A multi-phase deep CNN based mitosis detection framework for breast cancer histopathological images.

The mitotic activity index is a key prognostic measure in tumour grading. Microscopy based detection of mitotic nuclei is a significant overhead and necessitates automation. This work proposes deep CNN based multi-phase mitosis detection framework "MP-MitDet" for mitotic nuclei identification in breast cancer histopathological images. The workflow constitutes: (1) label-refiner, (2) tissue-level mitotic region selection, (3) blob analysis, and (4) cell-level refinement. We developed an automatic label-refiner to represent weak labels with semi-sematic information for training of deep CNNs. A deep instance-based detection and segmentation model is used to explore probable mitotic regions on tissue patches. More probable regions are screened based on blob area and then analysed at cell-level by developing a custom CNN classifier "MitosRes-CNN" to filter false mitoses. The performance of the proposed "MitosRes-CNN" is compared with the state-of-the-art CNNs that are adapted to cell-level discrimination through cross-domain transfer learning and by adding task-specific layers. The performance of the proposed framework shows good discrimination ability in terms of F-score (0.75), recall (0.76), precision (0.71) and area under the precision-recall curve (0.78) on challenging TUPAC16 dataset. Promising results suggest good generalization of the proposed framework that can learn characteristic features from heterogenous mitotic nuclei.

Depth of field multiplexing in microscopy.

We demonstrate "depth of field multiplexing" by a high resolution spatial light modulator (SLM) in a Fourier plane in the imaging path of a standard microscope. This approach provides simultaneous imaging of different focal planes in a sample with only a single camera exposure. The phase mask on the SLM corresponds to a set of superposed multi-focal off-axis Fresnel lenses, which sharply image different focal planes of the object to non-overlapping adjacent sections of the camera chip. Depth of field multiplexing allows to record motion in a three dimensional sample volume in real-time, which is exemplarily demonstrated for cytoplasmic streaming in plant cells and rapidly swimming protozoa.

Quantitative SLM-based Differential Interference Contrast imaging.

We describe the implementation of quantitative Differential Interference Contrast (DIC) Microscopy using a spatial light modulator (SLM) as a flexible Fourier filter in the optical path. The experimental arrangement allows for the all-electronic acquisition of multiple phase shifted DIC-images at video rates which are analyzed to yield the optical path length variation of the sample. The resolution of the technique is analyzed by retrieving the phase profiles of polystyrene spheres in immersion oil, and the method is then applied for quantitative imaging of biological samples. By reprogramming the diffractive structure displayed at the SLM it is possible to record the whole set of phase shifted DIC images simultaneously in different areas of the same camera chip. This allows for quantitative snap-shot imaging of a sample, which has applications for the investigation of dynamic processes.

Potentiality of using front face fluorescence spectroscopy for quantitative analysis of cow milk adulteration in buffalo milk.

In current study, synchronous front-face fluorescence spectroscopy together with partial least squares regression (PLSR) is used to predict the adulteration of cow and buffalo milk quantitatively. Fresh (unprocessed milk) samples of cow and buffalo were collected from local dairy farms. Fluorescence emission from milk samples mixed in different concentrations, show intensity variations at wavelengths 370-380 nm, 410 nm, 442 nm and 520-560 nm. Among them, the emissions at band position of 442 nm and 525 nm are highly selective between the two species and could help in finding adulteration of cow milk in buffalo milk and vice versa. The emissions at these wavelength positions correspond to fat-soluble vitamin-A as well as ß-carotene. PLS regression is used as a statistical prediction model, which is developed by training with the emission spectra of milk samples having known level of adulterations. The developed model predicts the unknown level of adulterations by means of their spectral data. The goodness of the model is determined by the correlation coefficient R-square (r2) value, which in our case is 0.99. Furthermore, the model root mean square error in cross validation (RMSECV) and in prediction (RMSECP) remains 1.16 and 6.24 respectively. This approach can effectively be applied to determine milk adulterations among other species as well as in detecting external agents (fraudulent) added into milk and other dairy products by further studies.

Cost effective and efficient screening of tuberculosis disease with Raman spectroscopy and machine learning algorithms.

The current study presents Raman Spectroscopy (RS) accompanied by machine learning algorithms based on Principle Component Analysis (PCA) and Hierarchical Cluster Analysis (HCA) for analysis of tuberculosis (TB). TB positive (diseased), TB negative (cured) and control (healthy) serum samples are considered for inter and intra comparative analysis. Raman spectral differences observed between both TB group and control samples spectra attributed probably to the changes in biomolecules like higher lactate concentration, lowering level of ß-carotene and amide-I band of protein in TB patient's blood samples. Inter comparison between control and TB positive sera samples shows prominent decrease in three extremely intense Raman peaks associated to ß-carotene concentration. Noteworthy spectral differences are also observed among TB positive and TB negative sera samples. The comparison of these Raman results clearly indicate that the blood composition of TB negative patients still showing irregularities in some important elements. Moreover, the Raman spectral differences observed in the data of the control and diseased samples are further highlighted with the help of the machine learning algorithms. In general, a fine correlation has been observed between PCA score plot as well as HCA dendogram with the original Raman findings. Further investigation of such noticeable differences could help in understandings regarding the existing threshold levels. Moreover in future, it can contribute a lot towards the development of new, modified and more effective screening options.

Demonstrating the application of Raman spectroscopy together with chemometric technique for screening of asthma disease.

Medical biophotonic tools provide new sources of diagnostic information regarding the state of human health that are used in managing patient care. In our current study, Raman spectroscopy, together with the chemometric technique, has successfully been demonstrated for the screening of asthma disease. Raman spectra of sera samples from asthmatic patients as well as healthy (control) volunteers have been recorded at 532 nm excitation. In healthy sera, three highly reproducible Raman peaks assigned to ß-carotene have been detected. Their sensitive detection is facilitated due to the resonance Raman effect. In contrast, in asthmatic patients sera, the peaks assigned to ß-carotene are either diminished or suppressed accompanied by other new Raman peaks. These new peaks most probably arise due to an elevated level of proteins, which could be used to identify/differentiate between asthma and non-asthma samples. Furthermore, a partial least squares discrimination analysis (PLS-DA) model was developed and applied on the Raman spectra of diseased as well as healthy samples, which successfully classified them. The correlation coefficient (r2) of the model was determined as 0.965. Similarly, the root mean square errors in cross-validation (RMSECV) and in the prediction (RMSECP) are 0.09 and 0.25, respectively. PLS-DA has the potential to be incorporated in a microcontroller's code attached with a hand-held Raman spectrometer for screening purposes in asthma, which is a disease of great concern for the clinicians, especially in children.

Optical screening of hepatitis-B infected blood sera using optical technique and neural network classifier.

In this study we demonstrate the analysis of biochemical changes in the human blood sera infected with Hepatitis B virus (HBV) using Raman spectroscopy. In total, 120 diseased blood samples and 170 healthy blood samples, collected from Pakistan Atomic Energy Commission (PAEC) general hospital, were analyzed. Spectra from each sample of both groups were collected in the spectral range 400-1700 cm-1. Careful spectral analyses demonstrated significant spectral variations (p < 0.0001) in the HBV infected individuals as compared to the normal ones. The spectral variations presumably occur because of the variations in the concentration of important biomolecules. Variations in spectral signatures were further exploited by using a neural network classifier towards machine-assisted classification of the two groups. Evaluation metrics of the classifier showed the diagnostic accuracy of (0.993), sensitivity ( = 0.992), specificity ( = 0.994), positive predictive value ( = 0.992) and negative predictive value ( = 0.994). The observed variations in the molecular concentration may be important markers of the hepatic performance and can be used in the diagnosis and machine-assisted classification of HBV infection.

A comparative study of machine learning classifiers for risk prediction of asthma disease.

Asthma is a chronic disease characterized by wheezing, chest tightening and difficulty in breathing due to inflammation of lung airways. Early risk prediction of asthma is crucial for proper and effective management. This study presents the use of machine learning approach for risk prediction of asthma by evaluating Raman spectral variations between asthmatic as well as healthy sera samples. Specifically, Raman spectra from 150 asthma and 52 healthy control blood sera samples were acquired. Spectral analyses illustrated significant spectral variations (p < 0.0001) in the asthmatic samples when compared with healthy sera. The existing spectral differences were further exploited by using artificial neural network (ANN) along with support vector machine (SVM) and random forest (RF) algorithms towards machine-assisted classification of the two groups. Quantitative comparison of the evaluation metrics of the classification algorithms showed superior performance of SVM model. Our results indicate that Raman spectroscopy in tandem with SVM can be used in the diagnosis and machine-assisted classification of asthma patients with promising accuracy.

Raman spectroscopy combined with a support vector machine for differentiating between feeding male and female infants mother's milk.

This study presents differentiation in milk samples of mother's feeding male and female infants using Raman spectroscopy combined with a support vector machine (SVM). Major differences have been observed in the Raman spectra of both types of milk based on their chemical compositions. Overall, it has been found that milk samples of mother's having a female infant are richer in fatty acids, phospholipids, and tryptophan. In contrast, milk samples of mother's having a male infant contain more carotenoids and saccharides. Principal component analysis and SVM further highlighted the differences between the two groups on the basis of differentiating features obtained from their Raman spectra. The SVM model with two different kernels, i.e. polynomial kernel function (order-2) and Gaussian radial basis function (RBF sigma-2), are used for gender based milk differentiations. The performance of the proposed model in terms of accuracy, precision, sensitivity, and specificity using the polynomial kernel function of order-2 have been found to be 86%, 88%, 85% and 88%, respectively.

Validation of Fluorescence Spectroscopy to Detect Adulteration of Edible Oil in Extra Virgin Olive Oil (EVOO) by Applying Chemometrics.

Due to high price and nutritional values of extra virgin olive oil (EVOO), it is vulnerable to adulteration internationally. Refined oil or other vegetable oils are commonly blended with EVOO and to unmask such fraud, quick, and reliable technique needs to be standardized and developed. Therefore, in this study, adulteration of edible oil (sunflower oil) is made with pure EVOO and analyzed using fluorescence spectroscopy (excitation wavelength at 350 nm) in conjunction with principal component analysis (PCA) and partial least squares (PLS) regression. Fluorescent spectra contain fingerprints of chlorophyll and carotenoids that are characteristics of EVOO and differentiated it from sunflower oil. A broad intense hump corresponding to conjugated hydroperoxides is seen in sunflower oil in the range of 441-489 nm with the maximum at 469 nm whereas pure EVOO has low intensity doublet peaks in this region at 441 nm and 469 nm. Visible changes in spectra are observed in adulterated EVOO by increasing the concentration of sunflower oil, with an increase in doublet peak and correspondingly decrease in chlorophyll peak intensity. Principal component analysis showed a distinct clustering of adulterated samples of different concentrations. Subsequently, the PLS regression model was best fitted over the complete data set on the basis of coefficient of determination (R2), standard error of calibration (SEC), and standard error of prediction (SEP) of values 0.99, 0.617, and 0.623 respectively. In addition to adulterant, test samples and imported commercial brands of EVOO were also used for prediction and validation of the models. Fluorescence spectroscopy combined with chemometrics showed its robustness to identify and quantify the specified adulterant in pure EVOO.

Analysis of hepatitis C infection using Raman spectroscopy and proximity based classification in the transformed domain.

This work presents a diagnostic system for the hepatitis C infection using Raman spectroscopy and proximity based classification. The proposed method exploits transformed Raman spectra using the proximity based machine learning technique and is denoted as RS-PCA-Prox. First, Raman spectral data is baseline corrected by subtracting noise and low intensity background. After this, a feature transformation of Raman spectra is adopted, not only to reduce the feature's dimensionality but also to learn different deviations in Raman shifts. The proposed RS-PCA-Prox shows significant diagnostic power in terms of accuracy, sensitivity, and specificity as 95%, 0.97 and 0.94 in PCA based transformed domain. The comparison of the RS-PCA-Prox with linear and ensemble based classifiers shows that proximity based classification performs better for the discrimination of HCV infected individuals and is able to differentiate the infected individuals from normal ones on the basis of molecular spectral information. Furthermore, it is observed that characteristic spectral changes are due to variation in the intensity of lectin, chitin, lipids, ammonia and viral protein as a consequence of the HCV infection.

Analysis of tuberculosis disease through Raman spectroscopy and machine learning.

We present the effectiveness of Raman spectroscopy (RS) in combination with machine learning for screening and analysis of blood sera collected from tuberculosis patients. Blood samples of 60 patients have confirmed active pulmonary tuberculosis and 14 samples of healthy age matched control were used in the current study. Spectra from entire sera samples were acquired using 785 nm laser Raman system. Support Vector Machine (SVM) together with Principal Component Analysis (PCA) has been used for highlighting variations spectral intensities between healthy and pathological samples. SVM model using Gaussian radial basis is able to discriminate between healthy and diseased patients based on the differences in the concentration of essential biomolecules such as lactate, ß-carotene, and amide-I. Diagnostic accuracy of 92%, with precision, specificity and sensitivity of 95%, 98% and 81%, respectively, were achieved considering PC3 and PC4. Automatic analysis of the variations in the concentration of these molecules together with chemometrics can effectively be utilized for an early screening of tuberculosis through minimum invasion.

Analysis of hepatitis B virus infection in blood sera using Raman spectroscopy and machine learning.

This study presents the analysis of hepatitis B virus (HBV) infection in human blood serum using Raman spectroscopy combined with pattern recognition technique. In total, 119 confirmed samples of HBV infected sera, collected from Pakistan Atomic Energy Commission (PAEC) general hospital have been used for the current analysis. The differences between normal and HBV infected samples have been evaluated using support vector machine (SVM) algorithm. SVM model with two different kernels i.e. polynomial function and Gaussian radial basis function (RBF) have been investigated for the classification of normal blood sera from HBV infected sera based on Raman spectral features. Furthermore, the performance of the model with each kernel function has also been analyzed with two different implementations of optimization problem i.e. Quadratic programming and least square. 5-fold cross validation method has been used for the evaluation of the model. In the current study, best classification performance has been achieved for polynomial kernel of order-2. A diagnostic accuracy of about 98% with the precision of 97%, sensitivity of 100% and specificity of 95% has been achieved under these conditions.

Identification of cow and buffalo milk based on Beta carotene and vitamin-A concentration using fluorescence spectroscopy.

The current study presents the application of fluorescence spectroscopy for the identification of cow and buffalo milk based on ß-carotene and vitamin-A which is of prime importance from the nutritional point of view. All samples were collected from healthy animals of different breeds at the time of lactation in the vicinity of Islamabad, Pakistan. Cow and buffalo milk shows differences at fluorescence emission appeared at band position 382 nm, 440 nm, 505 nm and 525 nm both in classical geometry (right angle) setup as well as front face fluorescence setup. In front face fluorescence geometry, synchronous fluorescence emission shows clear differences at 410 nm and 440 nm between the milk samples of both these species. These fluorescence emissions correspond to fats, vitamin-A and ß-carotene. Principal Component Analysis (PCA) further highlighted these differences by showing clear separation between the two data sets on the basis of features obtained from their fluorescence emission spectra. These results indicate that classical geometry (fixed excitation wavelength) as well as front face (synchronous fluorescence emission) of cow and buffalo milk nutrients could be used as fingerprint from identification point of view. This same approach can effectively be used for the determination of adulterants in the milk and other dairy products.