Adaptive modeling and spectral estimation of nonstationary biomedical signals based on Kalman filtering

We describe an algorithm to estimate the instantaneous power spectral density (PSD) of nonstationary signals. The algorithm is based on a dual Kalman filter that adaptively generates an estimate of the autoregressive model parameters at each time instant. The algorithm exhibits superior PSD tracking performance in nonstationary signals than classical nonparametric methodologies, and does not assume local stationarity of the data. Furthermore, it provides better time-frequency resolution, and is robust to model mismatches. We demonstrate its usefulness by a sample application involving PSD estimation of intracranial pressure signals (ICP) from patients with traumatic brain injury (TBI).     

Atlas-based segmentation of neonatal brain MR images using a gray matter enhancing step

We propose an atlas-based segmentation framework for brain magnetic resonance images, specially designed to fit neonatal images, which pose additional difficulties due to the poor differentiation between the gray and white matter. The main contribution of our work consists of a gray matter enhancing step, which is applied to either the T1w or T2w modalities after standard preprocessing and alignment steps are carried out. Our enhancing step uses Hessian and box filters for the cortical gray matter and takes advantage of both local and non-local information for the subcortical gray matter. We consider four classes, and our framework has been evaluated using publicly available data from the NeoBrainS12 challenge.     

Presentation of a PCR-nuclease protection strategy for minimal residual disease monitoring in B-ALL

Methods for detecting residual malignant cells in patients suffering from lymphoid malignancies have neither been sufficiently sensitive nor easy to routinize, hampering a potential prediction of disease outcome. Taking advantage of clone-specific DNA sequences, generated during lymphocyte differentiation and the polymerase chain reaction, some strategies have been developed for several groups. Up to now the most specific and sensitive methodology, which consists of designing leukemia-specific oligonucleotides, requires sequencing of the complementary determining region III-DNA for each particular patient and is too laborious to be applied to each case for routine monitoring in most hospital laboratories. In an attempt to achieve an easy way to detect residual malignant cells in B lymphoproliferative diseases, we have used a new PCR-based approach, named here as PCR-nuclease protection assay, consisting of: (i) amplification of DNA segments corresponding to the complementarity determining region III of the immunoglobulin heavy chain genes from samples at disease diagnosis; (ii) isolation of the disease-specific single-stranded DNA; (iii) labeling of the single-stranded DNA to generate specific probes; (iv) hybridization to amplified DNA from samples corresponding to different disease phases; and (v) digestion with S1-nuclease. Using this approach, we could detect one malignant cell in a background of 10(5) healthy cells. The sensitivity and specificity of this approach compares with those of the above mentioned specific oligonucleotide strategy in detecting residual malignant B cells. Moreover, this strategy is much less tedious and could be used by most hospital laboratories.     

Mitigation of voltage sags, imbalances and harmonics in sensitive industrial loads by means of a series power line conditioner

Power quality is a very important topic nowadays. Sensitive industrial equipment should be protected against steady-state distortions and temporary transients in the distribution lines. Typical disturbances that affect the voltage waveform quality are harmonics, imbalances and sags. This paper presents a solution to protect sensitive loads against voltage disturbances that is based on a series power line conditioner. The goal of the proposed design is that the load does not suffer considerable input voltage variations. To achieve that, a Proportional-Resonant (PR) controller and a reference generator block based on a low-gain PLL are used, which avoids a sag detection block. Furthermore, a frequency adaptation loop is included in the PR controller, which provides a proper controller operation even with utility grid frequency deviations. Finally, a hardware in the loop (HIL) test rig is used to validate the system.     

Determination of anthropogenic and biogenic compounds on atmospheric aerosol collected in urban, biomass burning and forest areas in São Paulo, Brazil

This study was conducted at three sites of different characteristics in São Paulo State: São Paulo (SPA), Piracicaba (PRB) and Mata Atlântica Forest (MAT). PM₁₀, n-alkanes, pristane and phytane, PAHs, water-soluble ions and biomass burning tracers like levoglucosan and retene, were determined in quartz fiber filters. Samplings occurred on May 8th to August 8th, 2007 at the MAT site; on August 15th to 29th in 2007 and November 10th to 29th in 2008 at the PRB site and, March 13th to April 4th in 2007 and August 7th to 29th in 2008 at the SPA site.Aliphatic compounds emitted biogenically were less abundant at the urban sites than at the forest site, and its distribution showed the influence of tropical vascular plants. Air mass transport from biomass burning regions is likely to impact the sites with specific molecular markers.The concentrations of all species were variable and dependent of seasonal changes. In the most dry and polluted seasons, n-alkane and cation total concentrations were similar between the megacity and the biomass burning site. PAHs and inorganic ion abundances were higher at São Paulo than Piracicaba, yet, the site influenced by biomass burning seems to be the most impacted by the organic anion abundance in the atmosphere. Pristane and phytane confirm the contamination by petroleum residues at urban sites; at the MAT site, biological activity and long range transport of pollutants might influence the levels of pristane.     

Nanostructured disposable impedimetric sensors as tools for specific biomolecular interactions: sensitive recognition of concanavalin A

The development of sensors to detect specific weak biological interactions is still today a challenging topic. Characteristics of carbohydrate-protein (lectin) interactions include high specificity and low affinity. This work describes the development of nanostructured impedimetric sensors for the detection of concanavalin A (Con A) binding to immobilized thiolated carbohydrate derivatives (D-mannose or D-glucose) onto screen-printed carbon electrodes (SPCEs) modified with gold nanoparticles. Thiolated D-galactose derivative was employed as negative control to evaluate the selectivity of the proposed methodology. After binding the thiolated carbohydrate to the nanostructured SPCEs, different functionalized thiols were employed to form mixed self-assembled monolayers (SAM). Electrochemical impedance spectroscopy (EIS) was employed as a technique to evaluate the binding of Con A to selected carbohydrates through the increase of electron transfer resistance of the ferri/ferrocyanide redox probe at the differently SAM modified electrodes. Different variables of the assay protocol were studied in order to optimize the sensor performance. Selective Con A determinations were only achieved by the formation of mixed SAMs with adequate functionalized thiols. Important differences were obtained depending on the chain lengths and functional groups of these thiols. For the 3-mercapto-1-propanesulfonate mixed SAMs, the electron transfer resistance varied linearly with the Con A concentration in the 2.2-40.0 μg mL(-1) range for D-mannose and D-glucose modified sensors. Low detection limits (0.099 and 0.078 pmol) and good reproducibility (6.9 and 6.1%, n=10) were obtained for the D-glucose and D-mannose modified sensors, respectively, without any amplification strategy.     

Mid-infrared fiber-optic reflection spectroscopy (FORS) analysis of artists' alkyd paints on different supports

Mid-infrared Fourier transform fiber-optic reflection spectroscopy (mid-IR FORS) is a noninvasive and flexible spectroscopic technique. It is ideal in the art conservation field because of its portability for on-site and in situ analysis of art objects, analyses that require delicate handling, or analyses of objects that cannot be sampled. This paper studies the applicability of mid-IR FORS for the characterization of commercial artists' alkyd paints cast on different supports. As predicted, the quality of the spectra and intensity of characteristic peaks varied according to reflectivity, roughness, and materials used in the supports. The presence of organic binder was best identified by its carbonyl peak (the most intense) and CH(2) stretching peaks; however, this was not sufficient to distinguish between oil and alkyd binders. The differentiation and identification of alkyds and oils must rely on the unique fingerprint peaks. However, in some cases, the fingerprint peaks were difficult to interpret because of strong absorptions caused by inorganic paint fillers, often present in modern paint formulations, resulting in anomalous dispersion and reststrahlen distortions.     

Atom-selective imaging and mechanical atom manipulation using the non-contact atomic force microscope

We succeeded in distinguishing between oxygen and silicon atoms on an oxygen-adsorbed Si(111)7 x 7 surface, and also distinguished between silicon and tin atoms on Si(111)7 x 7-Sn intermixed and Si(111) square root(3) x square root(3)-Sn mosaic-phase surfaces using non-contact atomic force microscopy (NC-AFM) at room temperature. Atom species of individual atoms are specified from the number of each atom in NC-AFM images, the tip-sample distance dependence of NC-AFM images and/or the surface distribution of each atom. Further, based on the NC-AFM method but using soft nanoindentation, we achieved two kinds of mechanical vertical manipulation of individual atoms: removal of a selected Si adatom and deposition of a Si atom into a selected Si adatom vacancy on the Si(111)7 x 7 surface at 78 K. Here, we carefully and slowly indented a Si atom on top of a clean Si tip apex onto a predetermined Si adatom to remove the targeted Si adatom and onto a predetermined Si adatom vacancy to deposit a Si atom, i.e. to repair the targeted Si adatom vacancy. By combining the atom-selective imaging method with two kinds of mechanical atom manipulation, i.e. by picking up a selected atom species and by depositing that atom one by one at the assigned site, we hope to construct nanomaterials and nanodevices made from more than two kinds of atom species in the near future.