Maximum Entropy Processing Research Articles

Entropy and Gaussian Filter-Based Adaptive Active Contour for Segmentation of Skin Lesions.

Malignant melanoma is considered one of the deadliest skin diseases if ignored without treatment. The mortality rate caused by melanoma is more than two times that of other skin malignancy diseases. These facts encourage computer scientists to find automated methods to discover skin cancers. Nowadays, the analysis of skin images is widely used by assistant physicians to discover the first stage of the disease automatically. One of the challenges the computer science researchers faced when developing such a system is the un-clarity of the existing images, such as noise like shadows, low contrast, hairs, and specular reflections, which complicates detecting the skin lesions in that images. This paper proposes the solution to the problem mentioned earlier using the active contour method. Still, seed selection in the dynamic contour method has the main drawback of where it should start the segmentation process. This paper uses Gaussian filter-based maximum entropy and morphological processing methods to find automatic seed points for active contour. By incorporating this, it can segment the lesion from dermoscopic images automatically. Our proposed methodology tested quantitative and qualitative measures on standard dataset dermis and used to test the proposed method's reliability which shows encouraging results.

On a Robust MaxEnt Process Regression Model with Sample-Selection.

In a regression analysis, a sample-selection bias arises when a dependent variable is partially observed as a result of the sample selection. This study introduces a Maximum Entropy (MaxEnt) process regression model that assumes a MaxEnt prior distribution for its nonparametric regression function and finds that the MaxEnt process regression model includes the well-known Gaussian process regression (GPR) model as a special case. Then, this special MaxEnt process regression model, i.e., the GPR model, is generalized to obtain a robust sample-selection Gaussian process regression (RSGPR) model that deals with non-normal data in the sample selection. Various properties of the RSGPR model are established, including the stochastic representation, distributional hierarchy, and magnitude of the sample-selection bias. These properties are used in the paper to develop a hierarchical Bayesian methodology to estimate the model. This involves a simple and computationally feasible Markov chain Monte Carlo algorithm that avoids analytical or numerical derivatives of the log-likelihood function of the model. The performance of the RSGPR model in terms of the sample-selection bias correction, robustness to non-normality, and prediction, is demonstrated through results in simulations that attest to its good finite-sample performance.

Exactly scale-free scale-free networks

Many complex natural and physical systems exhibit patterns of interconnection that conform, approximately, to a network structure referred to as scale-free. Preferential attachment is one of many algorithms that have been introduced to model the growth and structure of scale-free networks. With so many different models of scale-free networks it is unclear what properties of scale-free networks are typical, and what properties are peculiarities of a particular growth or construction process. We propose a simple maximum entropy process which provides the best representation of what are typical properties of scale-free networks, and provides a standard against which real and algorithmically generated networks can be compared. As an example we consider preferential attachment and find that this particular growth model does not yield typical realizations of scale-free networks. In particular, the widely discussed “fragility” of scale-free networks is actually found to be due to the peculiar “hub-centric” structure of preferential attachment networks. We provide a method to generate or remove this latent hub-centric bias—thereby demonstrating exactly which features of preferential attachment networks are atypical of the broader class of scale-free networks. We are also able to statistically demonstrate whether real networks are typical realizations of scale-free networks, or networks with that particular degree distribution; using a new surrogate generation method for complex networks, exactly analogous the widely used surrogate tests of nonlinear time series analysis.

Performance tuning non-uniform sampling for sensitivity enhancement of signal-limited biological NMR

Non-uniform sampling (NUS) has been established as a route to obtaining true sensitivity enhancements when recording indirect dimensions of decaying signals in the same total experimental time as traditional uniform incrementation of the indirect evolution period. Theory and experiments have shown that NUS can yield up to two-fold improvements in the intrinsic signal-to-noise ratio (SNR) of each dimension, while even conservative protocols can yield 20-40% improvements in the intrinsic SNR of NMR data. Applications of biological NMR that can benefit from these improvements are emerging, and in this work we develop some practical aspects of applying NUS nD-NMR to studies that approach the traditional detection limit of nD-NMR spectroscopy. Conditions for obtaining high NUS sensitivity enhancements are considered here in the context of enabling (1)H,(15)N-HSQC experiments on natural abundance protein samples and (1)H,(13)C-HMBC experiments on a challenging natural product. Through systematic studies we arrive at more precise guidelines to contrast sensitivity enhancements with reduced line shape constraints, and report an alternative sampling density based on a quarter-wave sinusoidal distribution that returns the highest fidelity we have seen to date in line shapes obtained by maximum entropy processing of non-uniformly sampled data.

Real-time Acquisition of Three Dimensional NMR Spectra by Non-uniform Sampling and Maximum Entropy Processing

Of the experiments to shorten NMR measuring time by sparse sampling, non-uniform sampling (NUS) is advantageous. NUS miminizes systematic errors which arise due to the lack of samplings by randomization. In this study, I report the real-time acquisition of 3D NMR data using NUS and maximum-entropy (MaxEnt) data processing. The real-time acquisition combined with NUS can reduce NMR measuring time much more. Compared with multidimensional decomposition (MDD) method, which was originally suggested by Jaravine and Orekhov (JACS 2006, 13421-13426), MaxEnt is faster at least several times and more suitable for the realtime acquisition. The designed sampling schedule of current study makes all the spectra during acquisition have the comparable resulting resolutions by MaxEnt. Therefore, one can judge the quality of spectra easily by examining the intensities of peaks. I report two cases of 3D experiments as examples with the simulated subdataset from experimental data. In both cases, the spectra having good qualitie for data analysis could be obtained only with 3% of original data. Its corresponding NMR measuring time was 8 minutes for 3D HNCO of ubiquitin.

Disease-associated Sequence Variations Congregate in a Polyanion Recognition Patch on Human Factor H Revealed in Three-dimensional Structure

Mutations and polymorphisms in the regulator of complement activation, factor H, have been linked to atypical hemolytic uremic syndrome (aHUS), membranoproliferative glomerulonephritis, and age-related macular degeneration. Many aHUS patients carry mutations in the two C-terminal modules of factor H, which normally confer upon this abundant 155-kDa plasma glycoprotein its ability to selectively bind self-surfaces and prevent them from inappropriately triggering the complement cascade via the alternative pathway. In the current study, the three-dimensional solution structure of the C-terminal module pair of factor H has been determined. A binding site for a fully sulfated heparin-derived tetrasaccharide has been delineated using chemical shift mapping and the C3d/C3b-binding site inferred from sequence comparisons and computational docking. The resultant information allows assessment of the likely consequences of aHUS-associated amino acid substitutions in this critical region of factor H. It is striking that, excepting those likely to perturb the three-dimensional structure, aHUS-associated missense mutations congregate in the polyanion-binding site delineated in this study, thus potentially disrupting a vital mechanism for control of complement on self-surfaces in the microvasculature of the kidney. It is intriguing that a single nucleotide polymorphism predisposing to age-related macular degeneration occupies another region of factor H that harbors a polyanion-binding site.

Functional Insights from the Structure of the Multifunctional C345C Domain of C5 of Complement

The complement protein C5 initiates assembly of the membrane attack complex. This remarkable process results in lysis of target cells and is fundamental to mammalian defense against infection. The 150-amino acid residue domain at the C terminus of C5 (C5-C345C) is pivotal to C5 function. It interacts with enzymes that convert C5 to C5b, the first step in the assembly of the membrane attack complex; it also binds to the membrane attack complex components C6 and C7 with high affinity. Here a recombinant version of this C5-C345C domain is shown to adopt the oligosaccharide/oligonucleotide binding fold, with two helices packed against a five-stranded beta-barrel. The structure is compared with those from the netrin-like module family that have a similar fold. Residues critical to the interaction with C5-convertase cluster on a mobile, hydrophobic inter-strand loop that protrudes from the open face of the beta-barrel. The opposite, helix-dominated face of C5-C345C carries a pair of exposed hydrophobic side chains adjacent to a striking negatively charged patch, consistent with affinity for positively charged factor I modules in C6 and C7. Modeling of homologous domains from complement proteins C3 and C4, which do not participate in membrane attack complex assembly, suggests that this provisionally identified C6/C7-interacting face is indeed specific to C5.

Integrated imaging with a centrally obscured logarithmic asphere

We study an integrated computational imaging system which incorporates a centrally obscured logarithmic asphere lens and rapid maximum entropy processing. A careful study of imaging for a two-point object at various distances is presented to demonstrate diffraction limited resolution and a tenfold increase in the depth of field. Also the overall transfer function is defined and measured. This system is well suited to high-quality photography where one requires excellent resolution and good contrast with little or no artifacts.

Molecular time scale generalized Langevin equation theory and polynomial maximum entropy imaging of spectral densities

Molecular time scale generalized Langevin equation (MTGLE) theory is discussed as an approach to condensed phase dynamics. A polynomial maximum entropy (MaxEnt) process for imaging required MTGLE spectral densities based on knowledge of the moments of the spectral density is introduced. The process is based on the use of interpolation polynomials which serve both to image the spectral density as well as provide a numerical procedure to compute the inverse Hessian matrix in a Newton-type minimization. A default model is added to allow for the inclusion of additional information in forming the image. The polynomial MaxEnt imaging process is found to be a fast, numerically stable, computational procedure which produces images comparable in quality to images obtained by other imaging processes. The polynomial MaxEnt imaging process is examined in the context of imaging MTGLE bath spectral densities with special emphasis on a coupled linear chain model. Standard harmonic oscillator, Hamiltonian bath models such as Ohmic-exponential and Ohmic-Gaussian are shown to possess regions of parameter space for which the MTGLE adiabatic frequency is imaginary. When the adiabatic frequency is zero, it is shown that imaging of the friction kernel is the best approach.

Gifa V. 4: A complete package for NMR data set processing

The Gifa program is designed for processing, displaying and analysing 1D, 2D and 3D NMR data sets. It has been constructed in a modular fashion, based on three independent modules: a set of commands that perform all the basic processing operations such as apodisation functions, a complete set of Fourier Transforms, phasing and baseline correction, peak-picking and line fitting, linear prediction and maximum entropy processing; a set of command language primitives that permit the execution of complex macro commands; and a set of graphic commands that permit to build a complete graphic user interface, allowing the user to interact easily with the program. We have tried to create a versatile program that can be easily extended according to the user's requirements and that is adapted to a novice as well as an experienced user. The program runs on any UNIX computer, with or without graphic display, in interactive or batch mode.

Analysis of complex protein and glycoprotein mixtures by electrospray ionization mass spectrometry with maximum entropy processing.

Analysis of complex protein and glycoprotein mixtures by electrospray ionization mass spectrometry with maximum entropy processing.

Maximum entropy processing of projections prior to SPET reconstruction

Maximum entropy processing of projections prior to SPET reconstruction

Deconvolution of planar scintigrams by maximum entropy

Planar scintigrams are deconvolved with a point spread function using the maximum entropy method with the aim of improving image quality. The technique requires the specification of several parameters. These are related to the level of noise present in the data and the one's a priori knowledge of the object imaged. The performance of the technique is tested for a wide range of these parameters using images of a Williams phantom in scattering material and a figure of merit, derived from the detectability of the smallest cold spot, is calculated. For close to optimal values of the parameters a factor of two improvement in the figure is found. A processed bone image shows improved contrast and resolution. Maximum entropy processing could be used to increase image quality or allow comparable image quality with reduced imaging time or patient dose.

Improvement in the characterization of minor drug metabolites from HPLC-NMR studies through the use of quantified maximum entropy processing of NMR spectra

Improvement in the characterization of minor drug metabolites from HPLC-NMR studies through the use of quantified maximum entropy processing of NMR spectra

On the maximum entropy problem with autocorrelations specified on a lattice

The maximum entropy process with autocorrelations specified on a finite lattice is identified as a Gaussian autoregressive process with a special structure in its coefficients. The autoregressive coefficients can be obtained by means of a fast algorithm. This result extends Burg's (1967) well-known maximum entropy theorem, where the autocorrelation is constrained for consecutive lags.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Suppression of artifacts in the phase-modulated rotating-frame imaging experiment using the maximum-entropy method

Phase-modulated rotating-frame imaging, a technique used. for the detection of localized metabolites in vivo, is usually implemented in a less than ideal manner because of the practical constraints imposed by the human or animal sample. As a consequence, the spectra obtained by Fourier transformation contain artifacts which distort or obscure the genuine spectral features. An attractive alternative is to use the maximum-entropy method to process the data, incorporating the time-domain response of the nuclear spins predicted by the Bloch equations. Here we demonstrate this approach using simulated data, data from phantoms, and data from the human liver and chest wall. We show that artifacts in the Fourier-transform spectra arising from off resonance effects, imperfect pulse angles, and truncation can be effectively avoided by maximum-entropy processing.

The maximum entropy method. Appearance and reality

It has often been asserted that the maximum entropy method can simultaneously suppress noise and enhance resolution in NMR spectra. We examine this claim using a Monte Carlo method to determine the accuracy and precision of line heights and integrals in spectra obtained by various forms of the maximum entropy method. It is shown that maximum entropy processing with lineshape deconvolution produces results that are roughly comparable to least-squares fitting of a model lineshape to the conventional Fourier transform spectrum.

Interstation surface wave attenuation by autoregressive deconvolution

A new technique for calculating interstation Green's functions and attenuation coefficients for seismic surface waves is presented. The interstation Green's function is evaluated from the autocorrelation functions of the seismograms, which are obtained from a maximum entropy process. Since a data-invariant time window is not used, the evaluated Green's functions gives reliable information on both the amplitude and the phase spectra of the system. This new technique is compared with other methods by applying them to both synthetic and real data from a path in the Canadian shield.

Generalized maximum entropy deconvolution of spectral segments

A maximum entropy Fourier spectral deconvolution (MEFSD) program capable of selecting segments of larger spectra is used to deconvolute overlapping peaks in experimental and synthetic test cases. Low signal-to-noise features in spectra which also have very large features (dynamic ranges 100:1, 1000:1, or more) can be selected for maximum entropy processing by performing calculations on a pseudo-FID constructed from the selected region. Results show efficient and powerful deconvolution with simultaneous noise suppression for spectral segments, even when relatively crowded. Quantitative tests of MEFSD on low signal-to-noise synthetic test spectra, modeled after experimental examples, showed accuracies which were at least as good, or better, in relative peak integrations, over conventional FFT processing followed by Lorentzian curve fitting, for segments from several different test cases with dynamic ranges of 1000:1 to 3:1.

An investigation of a maximum entropy method for the processing of 1H and 13C nmr spectra from glycosaminoglycan oligo- and poly-saccharides

Maximum Entropy (MEM) processing of 1H and 13C spectra from representative keratan sulphate, heparan sulphate and dermatan sulphate glycosaminoglycan polymers or oligomers is considered. The advantages and disadvantages of one specific approach, the MEM procedure, with regard to the accurate determinations of spin-spin couplings, chemical shifts and for resolution enhancements, are discussed. It is concluded that, providing the original spectrum has adequate signal-to-noise (S/N) ratio, useful resolution enhancement can be achieved without degradation of spectral noise, for the more accurate determinations of spin-spin couplings, chemical shifts and the unravelling of incompletely resolved structure. In situations of poor S/N ratio, noise uncertainties persist and artificial noise reduction by extra exponential smoothing followed by more severe MEM enhancement can lead to loss of spectral information.