Quantification of glycated hemoglobin and glucose in vivo using Raman spectroscopy and artificial neural networks.

Undiagnosed type 2 diabetes (T2D) remains a major public health concern. The global estimation of undiagnosed diabetes is about 46%, being this situation more critical in developing countries. Therefore, we proposed a non-invasive method to quantify glycated hemoglobin (HbA1c) and glucose in vivo. We developed a technique based on Raman spectroscopy, RReliefF as a feature selection method, and regression based on feed-forward artificial neural networks (FFNN). The spectra were obtained from the forearm, wrist, and index finger of 46 individuals. The use of FFNN allowed us to achieve an error in the predictive model of 0.69% for HbA1c and 30.12 mg/dL for glucose. Patients were classified according to HbA1c values into three categories: healthy, prediabetes, and T2D. The proposed method obtained a specificity and sensitivity of 87.50% and 80.77%, respectively. This work demonstrates the benefit of using artificial neural networks and feature selection techniques to enhance Raman spectra processing to determine glycated hemoglobin and glucose in patients with undiagnosed T2D.

Fabrication of Microbolometer Arrays Based on Polymorphous Silicon-Germanium.

This work reports the development of arrays of infrared sensors (microbolometers) using a hydrogenated polymorphous silicon-germanium alloy (pm-SixGe1-x:H). Basically, polymorphous semiconductors consist of an amorphous semiconductor matrix with embedded nanocrystals of about 2-3 nm. The pm-SixGe1-x:H alloy studied has a high temperature coefficient of resistance (TCR) of 4.08%/K and conductivity of 1.5 × 10-5 S∙cm-1. Deposition of thermosensing film was made by plasma-enhanced chemical vapor deposition (PECVD) at 200 °C, while the area of the devices is 50 × 50 µm2 with a fill factor of 81%. Finally, an array of 19 × 20 microbolometers was packaged for electrical characterization. Voltage responsivity values were obtained in the range of 4 × 104 V/W and detectivity around 2 × 107 cm∙Hz1/2/W with a polarization current of 70 µA at a chopper frequency of 30 Hz. A minimum value of 2 × 10-10 W/Hz1/2 noise equivalent power was obtained at room temperature. In addition, it was found that all the tested devices responded to incident infrared radiation, proving that the structure and mechanical stability are excellent.

Aspheric lens to increase the depth of focus.

For high-resolution optical systems, a long depth of focus is desirable. Unfortunately, resolution and depth of focus are inversely related. In this work, a novel lens is presented to produce long depth of focus beams, keeping the same resolution. The equations to perform the optical design of this kind of lenses and results are shown for a simple lens that can produce beams with a spot size of 2.9 µm over a range of 1.5 mm and for an achromatic doublet with a focus depth of 10 mm.

Wavefronts, caustics, and ronchigrams of a spherical wave reflected by a spherical mirror.

The aim of the present work is twofold: first we obtain analytical expressions for both the wavefronts and the caustic associated with the light rays reflected by a spherical mirror after being emitted by a point light source located at an arbitrary position in free space, and second, we describe, in detail, the structure of the ronchigrams when the grating or Ronchi ruling is placed at different relative positions to the caustic region and the point light source is located on and off the optical axis. We find that, in general, the caustic has two branches: one is a segment of a line, and the other is a two-dimensional surface. The wavefronts, at the caustic region, have self intersections and singularities. The ronchigrams exhibit closed-loop fringes when the grating is placed at the caustic region.

Optical study to identify and quantify capsaicin in optical window.

Capsaicin is an active compound in hot peppers. It has been studied for its health benefits for humans. Optical spectroscopy is an important tool for determining the optical properties or chemical composition of matter. The aim of this research is proposing an optical method to identify and quantify capsaicin in the visible range. To achieve this goal, we combined absorption and diffuse reflectance spectroscopy techniques to compute the extinction coefficient. Moreover, the concentration of the analytes was determined using the optical properties of capsaicin. Our method is a promising tool for developing a classification of capsaicin according to its percentage in chilies. The extinction coefficients are reported for 507nm and 663nm, which are the most significative. In addition, the coefficients to build the mathematical model for capsaicin are reported for Kubelka-Munk model. Finally, a comparison between capsaicin vs chilies spectra was obtained to identify spectral response. Diffuse reflectance signals allowed the identification of capsaicin and opened the possibility of this fast and easy to do method for classification and quantification of bioactive compounds.

Novel Assessment of Urinary Albumin Excretion in Type 2 Diabetes Patients by Raman Spectroscopy.

. Urinary albumin excretion remains the key biomarker to detect renal complications in type 2 diabetes. As diabetes epidemy increases, particularly in low-income countries, efficient and low-cost methods to measure urinary albumin are needed. In this pilot study, we evaluated the performance of Raman spectroscopy in the assessment of urinary albumin in patients with type 2 diabetes. The spectral Raman analysis of albumin was performed using artificial urine, at five concentrations of albumin and 24 h collection urine samples from ten patients with Type 2 Diabetes. The spectra were obtained after removing the background fluorescence and fitting Gaussian curves to spectral regions containing features of such metabolites. In the samples from patients with type 2 diabetes, we identified the presence of albumin in the peaks of the spectrum located at 663.07, 993.43, 1021.43, 1235.28, 1429.91 and 1633.91 cm-1. In artificial urine, there was an increase in the intensity of the Raman signal at 1450 cm-1, which corresponds to the increment of the concentrations of albumin. The highest concentration of albumin was located at 1630 cm-1. The capability of Raman spectroscopy for detection of small concentrations of urinary albumin suggests the feasibility of this method for the screening of type 2 diabetes renal complications.

Computation of the disk of least confusion for conic mirrors.

We use geometrical optics to compute, in an exact way and by using the third-order approximation, the disk of least confusion (DLC) or the best image produced by a conic reflector when the point source is located at any position on the optical axis. In the approximate case we obtain analytical formulas to compute the DLC. Furthermore, we apply our equations to particular examples to compare the exact and approximate results.

Métodos de clasificación para identificar lesiones en piel a partir de espectros de reflexión difusa / Classification methods to identify lesions in skin starting from spectra of diffuse reflectance

Con el objetivo de discriminar lesiones benignas y malignas en la piel humana a partir de espectros de reflexión difusa, se han analizado diferentes algoritmos de clasificación usando el software de aprendizaje automático y reconocimiento de patrones WEKA. Además, dada la alta dimensionalidad de la señal espectral, fue empleada una técnica selección de atributos para determinar las variables que aporten la mayor cantidad de información. Se probó la clasificación de la señal usando los algoritmos de máquinas de vectores de soporte, redes neuronales y bosques aleatorios, el desempeño fue evaluado usando el promedio de la k-fold cross-validation tomando en cuenta los porcentajes de instancias clasificadas correctamente, el índice kappa, el área bajo la curva ROC, la sensibilidad, y la especifidad. Finalmente se demuestra que el algoritmo de redes neuronales con los parámetros momentum y learning rate en 0,6 y 0,3 respectivamente, es el que mejor se adapta al problema de reconocimiento de patrones ya que clasifica correctamente al 89,89% de los casos.

In order to differentiate between benign and malignant lesions in the human skin using diffuse reflection spectra, different classification algorithms were tested using the WEKA data mining software. In addition, due to the high dimensionality of the spectral signal, an attribute selection technique was applied to determine the variables that contribute with more information. The spectral signal classification was tested using support vector machines, neural networks and random forests, their performance was measured using the k-fold cross-validation percentages of the Kappa statistic, area under the ROC curve, specificity and sensitivity. Finally it is shown that the one layer neural network with 6 neurons and the parameters momentum and learning rate in 0.6 and 0.3 respectively, is best suited to the problem of pattern recognition, achieving correctly classify 89.89% of the cases.