Functional adaptive controller for multivariable stochastic systems with dynamic structure of neural network

The article deals with a challenging problem of adaptive control design for multivariable stochastic systems with a functional uncertainty. Model of the system is based on multi‐layered perceptron neural networks where both the unknown parameters and the structure are found in real time without a necessity of any off‐line training process. The unknown parameters are estimated by a global estimation method, the Gaussian sum filter, and the structure of the neural network model is optimized by a proposed pruning method. The control law is based on a bicriterial approach to the suboptimal dual control. Two individual criteria are designed and used to introduce conflicting efforts between the estimation and control; probing and caution. A comparison of the proposed dual control and its alternative with an implementation of the pruning algorithm is shown in a numerical example. Copyright © 2011 John Wiley & Sons, Ltd.     

Digital PCR Panel for Sensitive Hematopoietic Chimerism Quantification after Allogeneic Stem Cell Transplantation

Accurate and sensitive determination of hematopoietic chimerism is a crucial diagnostic measure after allogeneic stem cell transplantation to monitor engraftment and potentially residual disease. Short tandem repeat (STR) amplification, the current “gold standard” for chimerism assessment facilitates reliable accuracy, but is hampered by its limited sensitivity (≥1%). Digital PCR (dPCR) has been shown to combine exact quantification and high reproducibility over a very wide measurement range with excellent sensitivity (routinely ≤0.1%) and thus represents a promising alternative to STR analysis. We here aimed at developing a whole panel of digital-PCR based assays for routine diagnostic. To this end, we tested suitability of 52 deletion/insertion polymorphisms (DIPs) for duplex analysis in combination with either a reference gene or a Y-chromosome specific PCR. Twenty-nine DIPs with high power of discrimination and good performance were identified, optimized and technically validated. We tested the newly established assays on retrospective patient samples that were in parallel also measured by STR amplification and found excellent correlation. Finally, a screening plate for initial genotyping with DIP-specific duplex dPCR assays was designed for convenient assay selection. In conclusion, we have established a comprehensive dPCR system for precise and high-sensitivity measurement of hematopoietic chimerism, which should be highly useful for clinical routine diagnostics.     

Future trends in data mining

Over recent years data mining has been establishing itself as one of the major disciplines in computer science with growing industrial impact. Undoubtedly, research in data mining will continue and even increase over coming decades. In this article, we sketch our vision of the future of data mining. Starting from the classic definition of “data mining”, we elaborate on topics that — in our opinion — will set trends in data mining.     

Cognitive agents — a procedural perspective relying on the predictability of Object-Action-Complexes (OACs)

Embodied cognition suggests that complex cognitive traits can only arise when agents have a body situated in the world. The aspects of embodiment and situatedness are being discussed here from the perspective of linear systems theory. This perspective treats bodies as dynamic, temporally variable entities, which can be extended (or curtailed) at their boundaries. We show how acting agents can, for example, actively extend their body for some time by incorporating predictably behaving parts of the world and how this affects the transfer functions. We suggest that primates have mastered this to a large degree increasingly splitting their world into predictable and unpredictable entities. We argue that temporary body extension may have been instrumental in paving the way for the development of higher cognitive complexity as it is reliably widening the cause-effect horizon about the actions of the agent. A first robot experiment is sketched to support these ideas.We continue discussing the concept of Object-Action Complexes (OACs) introduced by the European PACO-PLUS consortium to emphasize the notion that, for a cognitive agent, objects and actions are inseparably intertwined. In another robot experiment we devise a semi-supervised procedure using the OAC-concept to demonstrate how an agent can acquire knowledge about its world. Here the notion of predicting changes fundamentally underlies the implemented procedure and we try to show how this concept can be used to improve the robot’s inner model and behaviour. Hence, in this article we have tried to show how predictability can be used to augment the agent’s body and to acquire knowledge about the external world, possibly leading to more advanced cognitive traits.     

Soil surface illumination at micro-relief scale and soil BRDF data collected by a hyperspectral camera

The results of the paper draw attention to the fact that the hyperspectral image of soil surface at micro-relief scale may display variation in the soil spectral shape due to illumination conditions of the surface. The image of an extremely rough cultivated soil surface, very deeply ploughed, was obtained by a hyperspectral camera, in the range of 0.4–1.0 μm with 0.67–0.74 nm spectral resolution. It was found that the soil reflectance spectra of the studied surface, illuminated by the direct sunbeams, are clearly convex with distinct absorption features. Furthermore, the soil normalized reflectance spectra were used to distinguish the subtlety of the analysed shaded soil spectra shape. They show that depressions caused by the absorption features of O₂ and H₂O, contained in the atmosphere above directly illuminated soil fragments, transform into peaks, if the same soil is deeply shaded.     

Evaluation of electron beam welding technique for the integration of robust sensor elements into high-pressure automotive systems

For “on board” diagnosis purposes of the injected fuel quantity flow sensitive elements based on the thermo-resistive measurement principle were integrated into finished Common Rail injection nozzles of valve-covered orifice (VCO) or mini-sac hole (MSH) class. Electron-discharge machining as well as electron beam welding technique are key technologies for a reliable integration procedure. To demonstrate negligible influence on the hydraulic performance of the nozzle after modification an optical measurement set-up is used to record temporally resolved the propagation of the spray patterns ejected from the six injection holes simultaneously. From these investigations the impact on the structural distortion of the valve caused by the welding seam is proved as its position can be directly linked to any chances in spray performance of each individual injection hole. Reducing the energy input during the electron beam welding lowers substantially the asymmetry in spray patterns from hole to hole as the needle uncovers the six injection holes more symmetrically. Besides this important finding, the numerical calculations indicate that the implementation of the sensor chip slightly amplifies the asymmetry induced by the welding process due to an additional weakening of the nozzle body which is confirmed experimentally. Despite these challenges, however, it is demonstrated that appropriate parameters for the integration procedure can be found affecting the hydraulic performance negligible compared to the original state.     

GPU-based real-time approximation of the ablation zone for radiofrequency ablation

Percutaneous radiofrequency ablation (RFA) is becoming a standard minimally invasive clinical procedure for the treatment of liver tumors. However, planning the applicator placement such that the malignant tissue is completely destroyed, is a demanding task that requires considerable experience. In this work, we present a fast GPU-based real-time approximation of the ablation zone incorporating the cooling effect of liver vessels. Weighted distance fields of varying RF applicator types are derived from complex numerical simulations to allow a fast estimation of the ablation zone. Furthermore, the heat-sink effect of the cooling blood flow close to the applicator's electrode is estimated by means of a preprocessed thermal equilibrium representation of the liver parenchyma and blood vessels. Utilizing the graphics card, the weighted distance field incorporating the cooling blood flow is calculated using a modular shader framework, which facilitates the real-time visualization of the ablation zone in projected slice views and in volume rendering. The proposed methods are integrated in our software assistant prototype for planning RFA therapy. The software allows the physician to interactively place virtual RF applicator models. The real-time visualization of the corresponding approximated ablation zone facilitates interactive evaluation of the tumor coverage in order to optimize the applicator's placement such that all cancer cells are destroyed by the ablation.     

Local image structures and optic flow estimation

Different kinds of local image structures (such as homogeneous, edge-like and junction-like patches) can be distinguished by the intrinsic dimensionality of the local signals. Intrinsic dimensionality makes use of variance from a point and a line in spectral representation of the signal in order to classify it as homogeneous, edge-like or junction-like. The concept of intrinsic dimensionality has been mostly exercised using discrete formulations; however, recent work has introduced a continuous definition. The current study analyzes the distribution of local patches in natural images according to this continuous understanding of intrinsic dimensionality. This distribution reveals specific patterns than can be also associated to local image structures established in computer vision and which can be related to orientation and optic flow features. In particular, we link quantitative and qualitative properties of optic-flow error estimates to these patterns. In this way, we also introduce a new tool for better analysis of optic flow algorithms.     

Shape from Shading Using Probability Functions and Belief Propagation

Shape-from-shading (SFS) aims to reconstruct the three-dimensional shape of an object from a single shaded image. This article proposes an improved framework based on belief propagation for computing SFS. The implementation of the well-known brightness, integrability and smoothness constraints inside this framework is shown. We implement the constraints as probability density functions. For example, the brightness constraint is a two-dimensional probability density function that relates all possible surface gradients at a pixel to their probability given the pixel intensity. A straightforward extension of the framework to photometric stereo is presented, where multiple images of the same scene taken under different lighting conditions are available. The results are promising, especially since the solution is obtained by iteratively applying simple operations on a regular grid of points. The presented framework therefore can be implemented in parallel and is a reasonably likely biological scheme.