Prediction- and simulation-error based perceptron training: Solution space analysis and a novel combined training scheme

Previous papers have noted the difficulty in obtaining neural models which are stable under simulation when trained using prediction-error-based methods. Here the differences between series-parallel and parallel identification structures for training neural models are investigated. The effect of the error surface shape on training convergence and simulation performance is analysed using a standard algorithm operating in both training modes. A combined series-parallel/parallel training scheme is proposed, aiming to provide a more effective means of obtaining accurate neural simulation models. Simulation examples show the combined scheme is advantageous in circumstances where the solution space is known or suspected to be complex. (c) 2006 Elsevier B.V. All rights reserved.

Ligand and structure-based methodologies for the prediction of the activity of G protein-coupled receptor ligands

Accurate in silico models for the quantitative prediction of the activity of G protein-coupled receptor (GPCR) ligands would greatly facilitate the process of drug discovery and development. Several methodologies have been developed based on the properties of the ligands, the direct study of the receptor-ligand interactions, or a combination of both approaches. Ligand-based three-dimensional quantitative structure-activity relationships (3D-QSAR) techniques, not requiring knowledge of the receptor structure, have been historically the first to be applied to the prediction of the activity of GPCR ligands. They are generally endowed with robustness and good ranking ability; however they are highly dependent on training sets. Structure-based techniques generally do not provide the level of accuracy necessary to yield meaningful rankings when applied to GPCR homology models. However, they are essentially independent from training sets and have a sufficient level of accuracy to allow an effective discrimination between binders and nonbinders, thus qualifying as viable lead discovery tools. The combination of ligand and structure-based methodologies in the form of receptor-based 3D-QSAR and ligand and structure-based consensus models results in robust and accurate quantitative predictions. The contribution of the structure-based component to these combined approaches is expected to become more substantial and effective in the future, as more sophisticated scoring functions are developed and more detailed structural information on GPCRs is gathered.

Robust face recognition with partial occlusion, illumination variation and limited training data by optimal feature selection

This study investigates face recognition with partial occlusion, illumination variation and their combination, assuming no prior information about the mismatch, and limited training data for each person. The authors extend their previous posterior union model (PUM) to give a new method capable of dealing with all these problems. PUM is an approach for selecting the optimal local image features for recognition to improve robustness to partial occlusion. The extension is in two stages. First, authors extend PUM from a probability-based formulation to a similarity-based formulation, so that it operates with as little as one single training sample to offer robustness to partial occlusion. Second, they extend this new formulation to make it robust to illumination variation, and to combined illumination variation and partial occlusion, by a novel combination of multicondition relighting and optimal feature selection. To evaluate the new methods, a number of databases with various simulated and realistic occlusion/illumination mismatches have been used. The results have demonstrated the improved robustness of the new methods.

Bimodal biometric verification based on face and lips

In this paper, we present a novel approach to person verification by fusing face and lip features. Specifically, the face is modeled by the discriminative common vector and the discrete wavelet transform. Our lip features are simple geometric features based on a lip contour, which can be interpreted as multiple spatial widths and heights from a center of mass. In order to combine these features, we consider two simple fusion strategies: data fusion before training and score fusion after training, working with two different face databases. Fusing them together boosts the performance to achieve an equal error rate as low as 0.4% and 0.28%, respectively, confirming that our approach of fusing lips and face is effective and promising.

Virtual Microscopy and Digital Pathology in Training and Education

Traditionally, education and training in pathology has been delivered using textbooks, glass slides and conventional microscopy. Over the last two decades, the number of web-based pathology resources has expanded dramatically with centralized pathological resources being delivered to many students simultaneously. Recently, whole slide imaging technology allows glass slides to be scanned and viewed on a computer screen via dedicated software. This technology is referred to as virtual microscopy and has created enormous opportunities in pathological training and education. Students are able to learn key histopathological skills, e.g. to identify areas of diagnostic relevance from an entire slide, via a web-based computer environment. Students no longer need to be in the same room as the slides. New human–computer interfaces are also being developed using more natural touch technology to enhance the manipulation of digitized slides. Several major initiatives are also underway introducing online competency and diagnostic decision analysis using virtual microscopy and have important future roles in accreditation and recertification. Finally, researchers are investigating how pathological decision-making is achieved using virtual microscopy and modern eyetracking devices. Virtual microscopy and digital pathology will continue to improve how pathology training and education is delivered.

The influence of evidence-based medicine training on decision-making in relation to over-the-counter medicines: a qualitative study

Objectives
To explore the role of evidence of effectiveness when making decisions about over-the-counter (OTC) medication and to ascertain whether evidence-based medicine training raised awareness in decision-making. Additionally, this work aimed to complement the findings of a previous study because all participants in this current study had received training in evidence-based medicine (unlike the previous participants).

Methods
Following ethical approval and an e-mailed invitation, face-to-face, semi-structured interviews were conducted with newly registered pharmacists (who had received training in evidence-based medicine as part of their MPharm degree) to discuss the role of evidence of effectiveness with OTC medicines. Interviews were recorded and transcribed verbatim. Following transcription, all data were entered into the NVivo software package (version 8). Data were coded and analysed using a constant comparison approach.

Key findings
Twenty-five pharmacists (7 males and 18 females; registered for less than 4 months) were recruited and all participated in the study. Their primary focus with OTC medicines was safety; sales of products (including those that lack evidence of effectiveness) were justified provided they did no harm. Meeting patient expectation was also an important consideration and often superseded evidence. Despite knowledge of the concept, and an awareness of ethical requirements, an evidence-based approach was not routinely implemented by these pharmacists. Pharmacists did not routinely utilize evidence-based resources when making decisions about OTC medicines and some felt uncomfortable discussing the evidence-base for OTC products with patients.

Conclusions
The evidence-based medicine training that these pharmacists received appeared to have limited influence on OTC decision-making. More work could be conducted to ensure that an evidence-based approach is routinely implemented in practice

Robust Multimodal Person Identification With Limited Training Data

This paper presents a novel method of audio-visual feature-level fusion for person identification where both the speech and facial modalities may be corrupted, and there is a lack of prior knowledge about the corruption. Furthermore, we assume there are limited amount of training data for each modality (e.g., a short training speech segment and a single training facial image for each person). A new multimodal feature representation and a modified cosine similarity are introduced to combine and compare bimodal features with limited training data, as well as vastly differing data rates and feature sizes. Optimal feature selection and multicondition training are used to reduce the mismatch between training and testing, thereby making the system robust to unknown bimodal corruption. Experiments have been carried out on a bimodal dataset created from the SPIDRE speaker recognition database and AR face recognition database with variable noise corruption of speech and occlusion in the face images. The system's speaker identification performance on the SPIDRE database, and facial identification performance on the AR database, is comparable with the literature. Combining both modalities using the new method of multimodal fusion leads to significantly improved accuracy over the unimodal systems, even when both modalities have been corrupted. The new method also shows improved identification accuracy compared with the bimodal systems based on multicondition model training or missing-feature decoding alone.