Scalp field maps and their characterization

The present chapter gives a comprehensive introduction into the display and quantitative characterization of scalp field data. After introducing the construction of scalp field maps, different interpolation methods, the effect of the recording reference and the computation of spatial derivatives are discussed. The arguments raised in this first part have important implications for resolving a potential ambiguity in the interpretation of differences of scalp field data. In the second part of the chapter different approaches for comparing scalp field data are described. All of these comparisons can be interpreted in terms of differences of intracerebral sources either in strength, or in location and orientation in a nonambiguous way. In the present chapter we only refer to scalp field potentials, but mapping also can be used to display other features, such as power or statistical values. However, the rules for comparing and interpreting scalp field potentials might not apply to such data. Generic form of scalp field data Electroencephalogram (EEG) and event-related potential (ERP) recordings consist of one value for each sample in time and for each electrode. The recorded EEG and ERP data thus represent a two-dimensional array, with one dimension corresponding to the variable “time” and the other dimension corresponding to the variable “space” or electrode. Table 2.1 shows ERP measurements over a brief time period. The ERP data (averaged over a group of healthy subjects) were recorded with 19 electrodes during a visual paradigm. The parietal midline Pz electrode has been used as the reference electrode.

Class-switzched B cells display response to therapeutic B-cell depletion in rheumatoid arthritis

Introduction Reconstitution of peripheral blood (PB) B cells after therapeutic depletion with the chimeric anti-CD20 antibody rituximab (RTX) mimics lymphatic ontogeny. In this situation, the repletion kinetics and migratory properties of distinct developmental B-cell stages and their correlation to disease activity might facilitate our understanding of innate and adaptive B-cell functions in rheumatoid arthritis (RA). Methods Thirty-five 'RTX-naïve' RA patients with active arthritis were treated after failure of tumour necrosis factor blockade in an open-label study with two infusions of 1,000 mg RTX. Prednisone dose was tapered according to clinical improvement from a median of 10 mg at baseline to 5 mg at 9 and 12 months. Conventional disease-modifying antirheumatic drugs were kept stable. Subsets of CD19+ B cells were assessed by flow cytometry according to their IgD and CD27 surface expression. Their absolute number and relative frequency in PB were followed every 3 months and were determined in parallel in synovial tissue (n = 3) or synovial fluid (n = 3) in the case of florid arthritis. Results Six of 35 patients fulfilled the European League Against Rheumatism criteria for moderate clinical response, and 19 others for good clinical response. All PB B-cell fractions decreased significantly in number (P < 0.001) after the first infusion. Disease activity developed independently of the total B-cell number. B-cell repopulation was dominated in quantity by CD27-IgD+ 'naïve' B cells. The low number of CD27+IgD- class-switched memory B cells (MemB) in the blood, together with sustained reduction of rheumatoid factor serum concentrations, correlated with good clinical response. Class-switched MemB were found accumulated in flaring joints. Conclusions The present data support the hypothesis that control of adaptive immune processes involving germinal centre-derived, antigen, and T-cell-dependently matured B cells is essential for successful RTX treatment.

Renal allografts with IF/TA display distinct expression profiles of metzincins and related genes

Chronic renal allograft injury is often reflected by interstitial fibrosis (IF) and tubular atrophy (TA) without evidence of specific etiology. In most instances, IF/TA remains an irreversible disorder, representing a major cause of long-term allograft loss. As members of the protease family metzincins and functionally related genes are involved in fibrotic and sclerotic processes of the extracellular matrix (ECM), we hypothesized their deregulation in IF/TA. Gene expression and protein level analyses using allograft biopsies with and without Banff'05 classified IF/TA illustrated their deregulation. Expression profiles of these genes differentiated IF/TA from Banff'05 classified Normal biopsies in three independent microarray studies and demonstrated histological progression of IF/TA I to III. Significant upregulation of matrix metalloprotease-7 (MMP-7) and thrombospondin-2 (THBS-2) in IF/TA biopsies and sera was revealed in two independent patient sets. Furthermore, elevated THBS-2, osteopontin (SPP1) and beta-catenin may play regulatory roles on MMP. Our findings further suggest that deregulated ECM remodeling and possibly epithelial to mesenchymal transition (EMT) are implicated in IF/TA of kidney transplants, and that metzincins and related genes play an important role in these processes. Profiling of these genes may be used to complement IF/TA diagnosis and to disclose IF/TA progression in kidney transplant recipients.

Evaluating retraining rules for semi-supervised learning in neural network based cursive word recognition

Training a system to recognize handwritten words is a task that requires a large amount of data with their correct transcription. However, the creation of such a training set, including the generation of the ground truth, is tedious and costly. One way of reducing the high cost of labeled training data acquisition is to exploit unlabeled data, which can be gathered easily. Making use of both labeled and unlabeled data is known as semi-supervised learning. One of the most general versions of semi-supervised learning is self-training, where a recognizer iteratively retrains itself on its own output on new, unlabeled data. In this paper we propose to apply semi-supervised learning, and in particular self-training, to the problem of cursive, handwritten word recognition. The special focus of the paper is on retraining rules that define what data are actually being used in the retraining phase. In a series of experiments it is shown that the performance of a neural network based recognizer can be significantly improved through the use of unlabeled data and self-training if appropriate retraining rules are applied.

Spatial audition in a static virtual environment : the role of auditory-visual interaction

The integration of the auditory modality in virtual reality environments is known to promote the sensations of immersion and presence. However it is also known from psychophysics studies that auditory-visual interaction obey to complex rules and that multisensory conflicts may disrupt the adhesion of the participant to the presented virtual scene. It is thus important to measure the accuracy of the auditory spatial cues reproduced by the auditory display and their consistency with the spatial visual cues. This study evaluates auditory localization performances under various unimodal and auditory-visual bimodal conditions in a virtual reality (VR) setup using a stereoscopic display and binaural reproduction over headphones in static conditions. The auditory localization performances observed in the present study are in line with those reported in real conditions, suggesting that VR gives rise to consistent auditory and visual spatial cues. These results validate the use of VR for future psychophysics experiments with auditory and visual stimuli. They also emphasize the importance of a spatially accurate auditory and visual rendering for VR setups.

Thin Capitalization Rules and Entrepreneurial Capital Structure Decisions

Tax planners often choose debt over equity financing. As this has led to increased corporate debt financing, many countries have introduced thin capitalization rules to secure their tax revenues. In a general capital structure model we analyze if thin capitalization rules affect dividend and financing decisions, and whether they can partially explain why corporations receive both debt and equity capital. We model the Belgian, German and Italian rules as examples. We find that the so-called Miller equilibrium and definite financing effects depend significantly on the underlying tax system. Further, our results are useful for the treasury to decide what thin capitalization type to implement.