Analysis of latent fingerprint deposits by infrared microspectroscopy

We report the use of infrared (IR) microspectroscopy for the analysis of fingerprint residues. The advantage of using an IR microscope lies in the ability to visualize and obtain spectra of individual particles and droplets that make up fingerprint ridge deposits at a spatial resolution of approximately 10 microm. Our initial results suggest that infrared microspectroscopy in reflection-absorption mode provides reproducible spectral analysis of fingerprint residue. Since infrared microspectroscopy is nondestructive to the sample, we will be able to study the changes in fingerprint ridge deposits as a function of time. The method holds promise for probing the difference between latent fingerprints of adults and children.     

Ultrasonic flaw classification in weldments using probabilistic neural networks

A probabilistic neural network is used here to classify flaws in weldments from their ultrasonic scattering signatures. It is shown that such a network is both simple to construct and fast to train. Probabilistic nets are also shown to be able to exhibit the high performance of other neural networks, such as feed forward nets trained via back-propagation, while possessing important advantages of speed, explicitness of their architecture, and physical meaning of their outputs. Probabilistic nets are also demonstrated to have performance equal to common statistical approaches, such as theK-nearest neighbor method, while retaining their unique advantages.     

Seismic vulnerability assessment of chemical plants through probabilistic neural networks

A chemical industrial plant represents a sensitive presence in a region and, in case of severe damage due to earthquake actions, its impact on social life and environment can be devastating. From the structural point of view, chemical plants count a number of recurrent elements, which are classifiable in a discrete set of typological families (towers, chimneys, cylindrical or spherical or prismatic tanks, pipes etc.). The final aim of this work is to outline a general procedure to be followed in order to assign a seismic vulnerability estimate to each element of the various typological families.In this paper, F.E. simulations allowed to create a training set, which has been used to train a probabilistic neural system. A sample application has concerned the seismic vulnerability of simple spherical tanks.     

Anisotropy studied by polarization-modulated fourier transform infrared reflection difference microspectroscopy

We investigated anisotropic optical behavior in solid-state materials using Fourier transform infrared reflection microspectroscopy in combination with polarization modulation. For a Ca1.8Sr0.2RuO4 crystal with an isotropic optical surface, we found the reflection difference to be very close to zero, independent of the azimuthal angle of the sample. A Ca1.4Sr0.6RuO4 crystal with an anisotropic optical surface, however, exhibited a large anisotropic optical response with a strong angular dependence following a sinusoidal behavior. Furthermore, we examined the spatial distribution of the reflection difference in Bi0.17Ca0.83MnO3+delta using infrared synchrotron radiation and could clearly distinguish microscopic anisotropic domains having different optical axes. These results demonstrate that our experimental scheme can be used as a powerful tool to spectrally and spatially resolve anisotropy of solid-state materials in the mid-infrared region.     

Theory of infrared microspectroscopy for intact fibers

Infrared microspectroscopy is widely used for the chemical analysis of small samples. In particular, spectral properties of small cylindrical samples are important in forensic analysis, understanding relationships between microstructure and mechanical properties in fibers or fiber composites, and development of cosmetics and drugs for hair. The diameters of the constituent cylinders are typically of the order of the central wavelength of light used to probe the sample. Hence, structure and material spectral response are coupled and recorded spectra are usually distorted to the extent of becoming useless for molecular identification. In this paper, we apply rigorous optical theory to predict the spectral distortions observed in IR microspectroscopic data of fibers. The theory is used, first, to compute the changes that are observed for cylinders of various dimensions under different instrument configurations when compared to the bulk spectrum from the same material. We provide a method to recover intrinsic material spectral response from fibers by correcting for distortion introduced by the cylindrical structure. The theory reported here should enable the routine use of IR microspectroscopy and imaging for the molecular analysis of cylindrical domains in complex materials.     

Use of neural networks for fitting of FE probabilistic scaling model parameters

The probabilistic crack approach, based on the Monte Carlo method, was recently developed for finite element analysis of concrete cracking and related size effects. In this approach the heterogeneity of the material is taken into account by considering the material properties (tensile strength, Young modulus, etc.) to vary spatially following a normal distribution. N samples of the vector of random variables are generated from a specific probability density function, and the N samples corresponding to a simulation are functions of the mean value and of the standard deviation that define the Gauss density function. The problem is that these statistical moments are not known, a priori, for the characteristic volume of the finite elements used in the analysis. The paper proposes an inverse finite element analysis using neural networks for the determination of the statistical distribution parameters (e.g., for a normal distribution, the mean and the standard deviation) from a given response of the structure (for instance, an average load-displacement curve). From FE-analysis of 4-point bending beam tests, it is shown that the backanalysis technique developed in this paper is a powerful tool to determine the probabilistic distribution functions at the material level from structural tests for material volumes which are generally not accessible to direct testing. This revised version was published online in July 2006 with corrections to the Cover Date.     

Optical interconnection for neural networks by use of a self-imaging function

A grating-lens combination unit is developed to form a scalingself-transform function that can self-image on scale. Then an arrayof many such grating-lens units is used for the optical interconnectionof a two-dimensional neural network, and experiments are carriedout. We find that our idea is feasible, the optical interconnectionsystem is simple, and optical adjustment is easy.     

Artificial neural networks for the automated detection of trichloroethylene by passive Fourier transform infrared spectrometry

Artificial neural networks are applied to the automated classification of trichloroethylene (TCE) signatures from passive Fourier transform infrared remote sensing interferogram data. Through the use of three data collection methods, a combination of laboratory and field data is acquired that allows the methodology to be evaluated under a variety of infrared background conditions and in the presence of potentially interfering compounds such as sulfur hexafluoride, methyl ethyl ketone, acetone, carbon tetrachloride, and ammonia. To maximize the computational efficiency of the network optimization, experimental design techniques are employed to develop a training protocol for the network that takes into account the relationships among five variables that are related either to the network architecture or to the training process. This protocol is implemented for the case of a back-propagation neural network (BNN) and is used to develop an optimized network for the detection of TCE. The classification performance of the network is assessed by comparing both TCE detection capabilities and false detection rates to similar classification results obtained with the technique of piecewise linear discriminant analysis (PLDA). When applied to prediction data withheld from the optimization of both the BNN and PLDA algorithms, the BNN method is observed to outperform PLDA overall, with TCE detection rates in excess of 99% and false detection rates less than 0.5%.     

Synchrotron infrared microspectroscopy characterization of heterogeneities in solid-state blended polymers

Immiscible polymers, including polystyrene (PS) and polyethylene terephthalate (PET), were blended in the solid state via mechanical attrition, the first step of a near net-shape manufacturing (NNSM) technique. Subsequent analysis via Fourier Transform Infrared Spectroscopy (FT-IR) with synchrotron radiation successfully distinguished between blend components. Characteristic absorbance peaks for each polymer allowed both qualitative and quantitative mapping within prepared samples. Reproducible area maps were created for 50/50 blends of polymethylmethacrylate (PMMA)/PET and PET/PS, which exhibited areas of macroscopic phase separation. Fluctuations in blend concentration were particularly evident for PS/PMMA. However, spatial resolution was shown to limit the detection of heterogeneities. Further modifications with the synchrotron IR apparatus will improve resolution and allow for the direct comparison of NNSM-processed and melt-blended polymers.     

Pose estimation from a two-dimensional view by use of composite correlation filters and neural networks

We present a technique to estimate the pose of a three-dimensional object from a two-dimensional view. We first compute the correlation between the unknown image and several synthetic-discriminant-function filters constructed with known views of the object. We consider both linear and nonlinear correlations. The filters are constructed in such a way that the obtained correlation values depend on the pose parameters. We show that this dependence is not perfectly linear, in particular for nonlinear correlation. Therefore we use a two-layer neural network to retrieve the pose parameters from the correlation values. We demonstrate the technique by simultaneously estimating the in-plane and out-of-plane orientations of an airplane within an 8-deg portion. We show that a nonlinear correlation is necessary to identify the object and also to estimate its pose. On the other hand, linear correlation is more accurate and more robust. A combination of linear and nonlinear correlations gives the best results.     

Estimation of optical constants of thin film by the use of artificial neural networks

A system for analyzing single-layer optical thin films has been formulated by the use of artificial neural networks. The training data sets stem from the computational results of the physical model of thin films, and they are used to train the artificial neural network, which, when done, can give values of film parameters in the millisecond time regime. The fast backpropagation algorithm is employed during training. The results of training are also given.     

In situ infrared microspectroscopy of approximately 850 million-year-old prokaryotic fossils

In situ infrared (IR) and Raman microspectroscopy have been conducted on Neoproterozoic, organic-walled microfossils (prokaryotic fossils) in doubly polished, petrographic thin sections in order to detect their spectral signatures. The microfossils are very well preserved and occur in black chert from the approximately 850 million-year-old Bitter Springs Formation, Northern Territory, Australia. Raman microspectroscopy on two species of microfossils, one a filament and the other a coccoid, shows disordered peaks (D peak, 1340 cm(-1)) and graphite peaks (G peak, 1600 cm(-1)), indicating that they consist of disordered carbonaceous materials. IR micro-mapping results of the filament reveal that the distributions of peak heights at 2920 cm(-1) (aliphatic CH(2)), 1585 cm(-1) (aromatic C-C), and 1370 cm(-1) (aliphatic CH(3)) match the shape of the filamentous microfossil. These results suggest that IR microspectroscopy can be used for in situ characterization of organic polar signatures that morphologically indicate microfossils embedded in chert by using doubly-polished rock (petrographic) thin section samples. Further, these methods can be applied to controversial microfossil-like structures to test their biogenic nature.     

Classification of Fourier transform infrared microscopic imaging data of human breast cells by cluster analysis and artificial neural networks

Cluster analysis and artificial neural networks (ANNs) are applied to the automated assessment of disease state in Fourier transform infrared microscopic imaging measurements of normal and carcinomatous immortalized human breast cell lines. K-means clustering is used to implement an automated algorithm for the assignment of pixels in the image to cell and non-cell categories. Cell pixels are subsequently classified into carcinoma and normal categories through the use of a feed-forward ANN computed with the Broyden-Fletcher-Goldfarb-Shanno training algorithm. Inputs to the ANN consist of principal component scores computed from Fourier filtered absorbance data. A grid search optimization procedure is used to identify the optimal network architecture and filter frequency response. Data from three images corresponding to normal cells, carcinoma cells, and a mixture of normal and carcinoma cells are used to build and test the classification methodology. A successful classifier is developed through this work, although differences in the spectral backgrounds between the three images are observed to complicate the classification problem. The robustness of the final classifier is improved through the use of a rejection threshold procedure to prevent classification of outlying pixels.     

Three stage cervical cancer classifier based on hybrid ensemble learning with modified binary PSO using pretrained neural networks

ABSTRACT

Cervical cancer is one of the major challenges in developing nations like India.In recent years, a lot of research has been done todetect cervical cancer at an early stage through the pap-smear test, human papillomavirus test (HPV), etc. In this study, we have proposed athree-stage cervical cancer classifier to classify cervical cells among normal and abnormal cells using a hybrid ensemble classifier based onfeatures extracted using pre-trained neural networks. Furthermore, this work extends to classify the cells among different levels of dysplastic mainly mild, moderate and severe. The accuracy achieved for 2-class classification among normal and abnormal cells is up to 100% while for 4-class classification among normal, mild, moderate and severe dysplastic cells is up to 98.91% and 99.12% for new and old Herlev university hospital datasets respectively.     

Imaging the material properties of bone specimens using reflection-based infrared microspectroscopy

Fourier transform infrared microspectroscopy (FTIRM) is a widely used method for mapping the material properties of bone and other mineralized tissues, including mineralization, crystallinity, carbonate substitution, and collagen cross-linking. This technique is traditionally performed in a transmission-based geometry, which requires the preparation of plastic-embedded thin sections, limiting its functionality. Here, we theoretically and empirically demonstrate the development of reflection-based FTIRM as an alternative to the widely adopted transmission-based FTIRM, which reduces specimen preparation time and broadens the range of specimens that can be imaged. In this study, mature mouse femurs were plastic-embedded and longitudinal sections were cut at a thickness of 4 μm for transmission-based FTIRM measurements. The remaining bone blocks were polished for specular reflectance-based FTIRM measurements on regions immediately adjacent to the transmission sections. Kramers-Kronig analysis of the reflectance data yielded the dielectric response from which the absorption coefficients were directly determined. The reflectance-derived absorbance was validated empirically using the transmission spectra from the thin sections. The spectral assignments for mineralization, carbonate substitution, and collagen cross-linking were indistinguishable in transmission and reflection geometries, while the stoichiometric/nonstoichiometric apatite crystallinity parameter shifted from 1032/1021 cm(-1) in transmission-based to 1035/1025 cm(-1) in reflection-based data. This theoretical demonstration and empirical validation of reflection-based FTIRM eliminates the need for thin sections of bone and more readily facilitates direct correlations with other methods such as nanoindentation and quantitative backscatter electron imaging (qBSE) from the same specimen. It provides a unique framework for correlating bone's material and mechanical properties.     

Fluorescence diagnostics of oil pollution in coastal marine waters by use of artificial neural networks

We discuss the problems with and the real possibilities of determining oil pollution in situ in coastal marine waters with fluorescence spectroscopy and of using artificial neural networks for data interpretation. In general, the fluorescence bands of oil and aquatic humic substance overlap. At oil concentrations in water from a few to tens of micrograms per liter, the intensity of oil fluorescence is considerably lower than that of humic substances at concentrations that typically are present in coastal waters. Therefore it is necessary to solve the problem of separating the small amount of oil fluorescence from the humic substance background in the spectrum. The problem is complicated because of possible interactions between the components and variations in the parameters of the fluorescence bands of humic substances and oil in water. Fluorescence spectra of seawater samples taken from coastal areas of the Black Sea, samples prepared in the laboratory, and numerically simulated spectra were processed with an artificial neural network. The results demonstrate the possibility of estimating oil concentrations with an accuracy of a few micrograms per liter in coastal waters also in cases in which the contribution from other organic compounds, primarily humic substances, to the fluorescence spectrum exceeds that of oil by 2 orders of magnitude and more.     

Modelling metabolic energy by neural networks

The apparent metabolic energy (EMA) of barley is modelled as a function of 12 easily obtainable analytical parameters by applying neural networks with the error back-propagation learning strategy. Kohonen maps and Ward's clustering technique have been used to define the objects for the training and test sets. The architecture of the neural network and the relevant parameters of error back-propagation learning have been optimised providing a RMS of 1.081 and a correlation coefficient (predicted versus found values) of 0.82. Contour maps of all variables including the output EMA value have been obtained by applying the counter-propagation learning strategy in a two-layer neural network. The responses yielded by the networks show that this method is capable of establishing a quantitative relationship between EMA and the original variables.     

Optical implementation of neural networks for face recognition by the use of nonlinear joint transform correlators

We describe a nonlinear joint transform correlator-based two-layer neural network that uses a supervised learning algorithm for real-time face recognition. The system is trained with a sequence of facial images and is able to classify an input face image in real time. Computer simulations and optical experimental results are presented. The processor can be manufactured into a compact low-cost optoelectronic system. The use of the nonlinear joint transform correlator provides good noise robustness and good image discrimination.