Showing papers by "Jagath C. Rajapakse published in 2006"

FPGA Implementations of Neural Networks

Abstract: The development of neural networks has now reached the stage where they are employed in a large variety of practical contexts. However, to date the majority of such implementations have been in software. While it is generally recognised that hardware implementations could, through performance advantages, greatly increase the use of neural networks, to date the relatively high cost of developing Application-Specific Integrated Circuits (ASICs) has meant that only a small number of hardware neurocomputers has gone beyond the research-prototype stage. The situation has now changed dramatically: with the appearance of large, dense, highly parallel FPGA circuits it has now become possible to envisage putting large-scale neural networks in hardware, to get high performance at low costs. This in turn makes it practical to develop hardware neural-computing devices for a wide range of applications, ranging from embedded devices in high-volume/low-cost consumer electronics to large-scale stand-alone neurocomputers. Not surprisingly, therefore, research in the area has recently rapidly increased, and even sharper growth can be expected in the next decade or so. Nevertheless, the many opportunities offered by FPGAs also come with many challenges, since most of the existing body of knowledge is based on ASICs (which are not as constrained as FPGAs). These challenges range from the choice of data representation, to the implementation of specialized functions, through to the realization of massively parallel neural networks; and accompanying these are important secondary issues, such as development tools and technology transfer. All these issues are currently being investigated by a large number of researchers, who start from different bases and proceed by different methods, in such a way that there is no systematic core knowledge to start from, evaluate alternatives, validate claims, and so forth. FPGA Implementations of Neural Networks aims to be a timely one that fill this gap in three ways: First, it will contain appropriate foundational material and therefore be appropriate for advanced students or researchers new to the field. Second, it will capture the state of the art, in both depth and breadth and therefore be useful researchers currently active in the field. Third, it will cover directions for future research, i.e. embryonic areas as well as more speculative ones.

Two-stage support vector regression approach for predicting accessible surface areas of amino acids.

Abstract: We address the problem of predicting solvent accessible surface area (ASA) of amino acid residues in protein sequences, without classifying them into buried and exposed types. A two-stage support vector regression (SVR) approach is proposed to predict real values of ASA from the position-specific scoring matrices generated from PSI-BLAST profiles. By adding SVR as the second stage to capture the influences on the ASA value of a residue by those of its neighbors, the two-stage SVR approach achieves improvements of mean absolute errors up to 3.3%, and correlation coefficients of 0.66, 0.68, and 0.67 on the Manesh dataset of 215 proteins, the Barton dataset of 502 nonhomologous proteins, and the Carugo dataset of 338 proteins, respectively, which are better than the scores published earlier on these datasets. A Web server for protein ASA prediction by using a two-stage SVR method has been developed and is available (http://birc.ntu.edu.sg/∼pas0186457/asa.html). Proteins 2006. © 2006 Wiley-Liss, Inc.

Contextual modeling of functional MR images with conditional random fields

Abstract: This paper presents a conditional random field (CRF) approach to fuse contextual dependencies in functional magnetic resonance imaging (fMRI) data for the detection of brain activation. The interactions among both activation (activated/inactive) labels and observed data of brain voxels are unified in a probabilistic framework based on the CRF, where the interaction strength can be adaptively adjusted in terms of the data similarity of neighboring sites. Compared to earlier detection methods, including statistical parametric mapping and Markov random field, the proposed method avoids the suppression of high frequency information and relaxes the strong assumption of conditional independence of observed data. Experimental results show that the proposed approach effectively integrates contextual constraints within the detection process and robustly detects brain activities from fMRI data

Exploratory analysis of brain connectivity with ICA

Abstract: Covariance-based methods of exploration of functional connectivity of the brain from functional magnetic resonance imaging (fMRI) experiments, such as principal component analysis (PCA) and structural equation modeling (SEM), require a priori knowledge such as an anatomical model to infer functional connectivity. In this research, a hybrid method, combining independent component analysis (ICA) and SEM, which is capable of deriving functional connectivity in an exploratory manner without the need of a prior model is introduced. The spatial ICA (SICA) derives independent neural systems or sources involved in task-related brain activation, while an automated method based on the SEM finds the structure of the connectivity among the elements in independent neural systems. Unlike second-order approaches used in earlier studies, the task-related neural systems derived from the ICA provide brain connectivity in the complete statistical sense. The use and efficacy of this approach is illustrated on two fMRI datasets obtained from a visual task and a language reading task.

NURBS-Based Visualization of Age-Related Diversity in Cortical Morphology

Abstract: Introduction In neurology, it is known that different motor and cognitive functions in humans are mapped on specific cortical surface patterns of the brain. The variability in these surface patterns between individuals is caused by local differences in the rate of angular deformations of cerebral tissues during brain development. It has been reported that age effects of the human brain result in significant changes in the cortical volume (Courchesne et al., 2000). Moreover, the locations of individual convolutions on the brain surface depend on local changes in cortical structure, with sulci occurring in relation to boundaries between cortical architectonic fields. Additionally, the importance of visualizing medical images, especially in 3D, for clinical applications involving the brain, is immense (Linney and Alusi, 1998). Therefore, a method to analyze and visualize age-related variability of surface patterns is useful in neuro-imaging studies to detect various functional disorders and behavioral changes between children and adults (Kolb, 1993). The investigation in this paper presents a new approach based on statistical analysis of curvature changes in deformation vectors of a NURBS surface to compute and visualize the variability associated with surface patterns of the human brain.

Protein-Protein Interface Residue Prediction with SVM Using Evolutionary Profiles and Accessible Surface Areas

Abstract: Knowledge of protein-protein interaction sites is vital to determine proteins' function and involvement in different pathways. Though a wide variety of methods has been proposed over the recent years in order to predict protein-protein interface residues, mainly based on single amino acid sequence inputs, each has its own drawbacks and limitations. We propose to use Support Vector Machines (SVM) to predict protein-protein interface residues by using the information derived from the position-specific scoring matrices (PSSMs) generated from PSI-BLAST profiles and accessible surface areas. The present approach achieved overall prediction accuracy of 73.8% for 77 individulal proteins collected from the Protein Data Bank, which is better than the previously reported accuracies.

Pattern recognition in bioinformatics: an introduction

Abstract: The information stored in DNA, a chain of four nucleotides (A, T, G, and C), is first converted to mRNA through the process of transcription and then converted to the functional form of life, proteins, through the process of translation. Only about 5% of the genome contains useful patterns of nucleotides, or genes, that code for proteins. The initiation of translation or transcription process is determined by the presence of specific patterns of DNA or RNA, or motifs. Research on detecting specific patterns of DNA sequences such as genes, protein coding regions, promoters, etc., leads to uncover functional aspects of cells. Comparative genomics focus on comparisons across the genomes to find conserved patterns over the evolution, which possess some functional significance. Construction of evolutionary trees is useful to know how genome and proteome are evolved over all species by ways of a complete library of motifs and genes.

Extraction of Fuzzy Features for Detecting Brain Activation from Functional MR Time-Series

Abstract: We propose methods to extract fuzzy features from fMR time-series in order to detect brain activation Five discriminating features are automatically extracted from fMRI using a sequence of temporal-sliding-windows A fuzzy model based on these features is first developed by gradient method training on a set of initial training data and then incrementally updated The resulting fuzzy activation maps are then combined to provide a measure of strength of activation for each voxel in human brain; a two-way thresholding scheme is introduced to determine actual activated voxels The method is tested on both synthetic and real fMRI datasets for functional activation detection, illustrating that it is less vulnerable to correlated noise and is able to adapt to different hemodynamic response functions across subjects through incremental learning

Graphical approach to weak motif recognition in noisy data sets

Abstract: Accurate recognition of motifs in biological sequences has become a central problem in computational biology. Though previous approaches have shown reasonable performances in detecting motifs having clear consensus, they are inapplicable to the recognition of weak motifs in noisy datasets, where only a fraction of the sequences may contain motif instances. This paper presents a graphical approach to deal with the real biological sequences, which are noisy in nature, and find potential weak motifs in the higher eukaryotic datasets. We examine our approach on synthetic datasets embedded with the degenerate motifs and show that it outperforms the earlier techniques. Moreover, the present approach is able to find the wet-lab proven motifs and other unreported significant consensus in real biological datasets.

Determination of the Relative Importance of Gene Function or Taxonomic Grouping to Codon Usage Bias Using Cluster Analysis and SVMs

Abstract: The codon usage patterns of 2,552 major histocompatibility complex (MHC) sequences from 33 primate species, and the consequent subsets of sequences obtained by removing species with most abundant sequences was observed. The correlation between function and species with regards to MHC codon usage patterns was analyzed using cluster analysis and Support Vector Machines (SVMs). The results show that gene function is the major factor, while species is the minor factor correlated to codon usage bias, but their interactions complicate the codon usage pattern. When the weight of the factor of species increases, the accuracy rate of classification dropped accordingly. The factors of gene function and species can be adopted as feature vectors in the field of gene classification and phylogenetic studies respectively. As the input of codon usage to the classifier is independent of sequence length and variance, our approach is useful when the sequences to be analyzed are of different lengths, a condition where classic homology-based approaches tend to be difficult. To focus on the phylogenetic features of the MHC sequences through codon usage analysis, we must try to minimize or even eliminate the influence of gene function.

Ultra-Fast fMRI Imaging with High-Fidelity Activation Map

Abstract: Functional Magnetic Resonance Imaging (fMRI) requires ultra-fast imaging in order to capture the on-going spatio-temporal dynamics of the cognitive task. We make use of correlations in both k-space and time, and thereby reconstruct the time series by acquiring only a fraction of the data, using an improved form of the well-known dynamic imaging technique k-t BLAST (Broad-use Linear Acquisition Speed-up Technique). k-t BLAST (κ-tB) works by unwrapping the aliased Fourier conjugate space of k-t ( y-f space). The unwrapping process makes use of an estimate of the true y-f space, obtained by acquiring a blurred unaliased version. In this paper, we propose two changes to the existing algorithm. Firstly, we improve the map estimate using generalized series reconstruction. The second change is to incorporate phase constraints from the training map. The proposed technique is compared with existing k-tB on visual stimulation fMRI data obtained on 5 volunteers. Results show that the proposed changes lead to gain in temporal resolution by as much as a factor of 6. Performance evaluation is carried out by comparing activation maps obtained using reconstructed images, against that obtained from the true images. We observe upto 10dB improvement in PSNR of activation maps. Besides, RMSE reduction on fMRI images, of about 10% averaged over the entire time series, with a peak improvement of 35% compared to the existing k-tB, averaged over 5 data sets, is also observed.

Effect of Diffusion Weighting and Number of Sensitizing Directions on Fiber Tracking in DTI

Abstract: Diffusion Tensor (DT) fiber tracking techniques offer significant potential for studying anatomical connectivity of human brain in vivo And the reliability and accuracy of fiber tracking results depend on the quality of estimated DT which is determined by parameters of image acquisition protocol The aim of this paper is to investigate what echo-planar image (EPI) acquisition parameters: the number of sensitizing directions K and diffusion weighting b-value gives the best estimation of DT and shorter scan time We carried out tracking on synthetic dataset that was artificially corrupted by various levels of Gaussian noise to study the effects of K and b-value on fiber tracking results, and to evaluate the quality of estimated DT It was found that when K value larger than 13 and b-value larger than 800 smm -2 best estimated DTs And further increments of K and b-value had no significant effect on quality of estimated DT

Extraction of fuzzy features for detecting brain activation from functional MR time-series

Abstract: We proposemethods to extract fuzzy features from fMR time-series in order to detect brain activation. Five discriminating features are automatically extracted from fMRI using a sequence of temporal-sliding-windows. A fuzzy model based on these features is first developed by gradientmethod training on a set of initial training data and then incrementally updated. The resulting fuzzy activation maps are then combined to provide a measure of strength of activation for each voxel in human brain; a two-way thresholding scheme is introduced to determine actual activated voxels. The method is tested on both synthetic and real fMRI datasets for functional activation detection, illustrating that it is less vulnerable to correlated noise and is able to adapt to different hemodynamic response functions across subjects through incremental learning.

Pattern Recognition in Bioinformatics: International Workshop, PRIB 2006, Hong Kong, China, August 20, 2006, Proceedings

Abstract: Pattern Recognition in Bioinformatics: An Introduction.- Pattern Recognition in Bioinformatics: An Introduction.- 1: Signal and Motif Detection Gene Selection.- Machine Learning Prediction of Amino Acid Patterns in Protein N-myristoylation.- A Profile HMM for Recognition of Hormone Response Elements.- Graphical Approach to Weak Motif Recognition in Noisy Data Sets.- Comparative Gene Prediction Based on Gene Structure Conservation.- Computational Identification of Short Initial Exons.- Pareto-Gamma Statistic Reveals Global Rescaling in Transcriptomes of Low and High Aggressive Breast Cancer Phenotypes.- Investigating the Class-Specific Relevance of Predictor Sets Obtained from DDP-Based Feature Selection Technique.- A New Maximum-Relevance Criterion for Significant Gene Selection.- 2: Models of DNA, RNA, and Protein Structures.- Spectral Graph Partitioning Analysis of In Vitro Synthesized RNA Structural Folding.- Predicting Secondary Structure of All-Helical Proteins Using Hidden Markov Support Vector Machines.- Prediction of Protein Subcellular Localizations Using Moment Descriptors and Support Vector Machine.- Using Permutation Patterns for Content-Based Phylogeny.- 3: Biological Databases and Imaging.- The Immune Epitope Database and Analysis Resource.- Intelligent Extraction Versus Advanced Query: Recognize Transcription Factors from Databases.- Incremental Maintenance of Biological Databases Using Association Rule Mining.- Blind Separation of Multichannel Biomedical Image Patterns by Non-negative Least-Correlated Component Analysis.- Image and Fractal Information Processing for Large-Scale Chemoinformatics, Genomics Analyses and Pattern Discovery.- Hybridization of Independent Component Analysis, Rough Sets, and Multi-Objective Evolutionary Algorithms for Classificatory Decomposition of Cortical Evoked Potentials.

Effect of diffusion weighting and number of sensitizing directions on fiber tracking in DTI

Abstract: Diffusion Tensor (DT) fiber tracking techniques offer significant potential for studying anatomical connectivity of human brain in vivo. And the reliability and accuracy of fiber tracking results depend on the quality of estimated DT which is determined by parameters of image acquisition protocol. The aim of this paper is to investigate what echo-planar image (EPI) acquisition parameters: the number of sensitizing directions K and diffusion weighting b-value gives the best estimation of DT and shorter scan time. We carried out tracking on synthetic dataset that was artificially corrupted by various levels of Gaussian noise to study the effects of K and b-value on fiber tracking results, and to evaluate the quality of estimated DT. It was found that when K value larger than 13 and b-value larger than 800 smm-2 best estimated DTs. And further increments of K and b-value had no significant effect on quality of estimated DT.

Ultra-Fast fMRI imaging with high-fidelity activation map

Abstract: Functional Magnetic Resonance Imaging (fMRI) requires ultra-fast imaging in order to capture the on-going spatio-temporal dynamics of the cognitive task We make use of correlations in both k-space and time, and thereby reconstruct the time series by acquiring only a fraction of the data, using an improved form of the well-known dynamic imaging technique k-t BLAST (Broad-use Linear Acquisition Speed-up Technique) k-t BLAST (k-tB) works by unwrapping the aliased Fourier conjugate space of k-t ( y-f space) The unwrapping process makes use of an estimate of the true y-f space, obtained by acquiring a blurred unaliased version In this paper, we propose two changes to the existing algorithm Firstly, we improve the map estimate using generalized series reconstruction The second change is to incorporate phase constraints from the training map The proposed technique is compared with existing k-tB on visual stimulation fMRI data obtained on 5 volunteers Results show that the proposed changes lead to gain in temporal resolution by as much as a factor of 6 Performance evaluation is carried out by comparing activation maps obtained using reconstructed images, against that obtained from the true images We observe upto 10dB improvement in PSNR of activation maps Besides, RMSE reduction on fMRI images, of about 10% averaged over the entire time series, with a peak improvement of 35% compared to the existing k-tB, averaged over 5 data sets, is also observed

Input encoding method for identifying transcription start sites in RNA polymerase II promoters by neural networks

Abstract: We present a novel approach to encode inputs to neural networks for the recognition of transcription start sites in RNA polymerase II promoter regions. The approach is based on Markov models that represent TATA-box and Inr transcription binding sites, characterizing a promoter. The Markovian parameters are used as inputs to three neural networks which learn potential distant relationships between the nucleotides at promoter regions. Such an approach allows for incorporating the biological contextual information of the promoter sites into neural network systems and implementing higher-order Markov models of the promoters. Our experiments on a human promoter data set, available at [19], showed an increased correlation coefficient rate of 0.69 on average, which is better than the earlier reported best rate of 0.65 by NNPP 2.1 method.