TutorialsComplex-valued Adaptive Signal Processing: Recent Advances in Optimization and Statistical Characterization
Group Independent Component Analysis for fMRI Data and ICA for Joint Inference of Imaging, Genetic, and ERP data
Note that the location for the tutorial is different of the location for the conference. See detail for access in the Venue page.
Complex-valued Adaptive Signal Processing: Recent Advances in Optimization and Statistical CharacterizationProfessor Tülay Adali
University of Maryland Baltimore County, USA
TÜLAY ADALI received the Ph.D. degree in electrical engineering from North Carolina State University, Raleigh, in 1992 and joined the faculty at the University of Maryland Baltimore County (UMBC), Baltimore, the same year. She is currently a Professor in the Department of Computer Science and Electrical Engineering at UMBC. She has held visiting positions at Technical University of Denmark, Lyngby, Denmark, Katholieke Universiteit, Leuven, Belgium, University of Campinas, Brazil, and École Supérieure de Physique et de Chimie Industrielles, Paris, France.
Prof. Adali assisted in the organization of a number of international conferences and workshops including the IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP), the IEEE International Workshop on Neural Networks for Signal Processing (NNSP), and the IEEE International Workshop on Machine Learning for Signal Processing (MLSP). She was the General Co-Chair, NNSP (2001-2003); Technical Chair, MLSP (2004-2008); Publicity Chair, ICASSP (2000 and 2005); and Publications Co-Chair, ICASSP 2008. She is currently the Program Co-Chair, 2009 MLSP and 2009 International Conference on Independent Component Analysis and Source Separation.
Prof. Adali chaired the IEEE SPS Machine Learning for Signal Processing Technical Committee (2003-2005); Member, SPS Conference Board (1998-2006); Member, Bio Imaging and Signal Processing Technical Committee (2004-2007); and Associate Editor, IEEE Transactions on Signal Processing (2003-2006). She is currently Chair of Technical Committee 14: Signal Analysis for Machine Intelligence of the International Association for Pattern Recognition; Member, Machine Learning for Signal Processing Technical Committee and an Associate Editor, IEEE Transactions on Biomedical Engineering, Elsevier Signal Processing Journal, and Journal of Signal Processing Systems for Signal, Image, and Video Technology.
Prof. Adali is a Fellow of the IEEE and the AIMBE. Her research interests are in the areas of statistical signal processing, machine learning for signal processing, biomedical data analysis (functional MRI, MRI, PET, CR, ECG, and EEG), bioinformatics, and signal processing for optical communications.
She is the recipient of a 1997 National Science Foundation (NSF) CAREER Award with more recent support from the National Institutes of Health, NSF, NASA, the US Army, and industry.
Complex-valued signals arise frequently in applications as diverse as communications, radar, and biomedicine, as most practical modulation formats are of complex type and applications such as radar and magnetic resonance imaging lead to data that are inherently complex valued. Complex-valued signal processing, however, has been long regarded as a simple extension of the real domain processing, and as a result, until recently most adaptive signal processing algorithms developed for the complex domain have failed to take full advantage of complex domain processing.
Two key issues that need to be addressed in this regard pertain to optimization in the complex domain and the statistical characterization of complex signals. In this tutorial, we address both issues and first introduce Wirtinger calculus that allows development of an efficient framework for optimization in the complex domain. We introduce the fundamental relationships between the real and the complex domains, and show how many real domain optimization algorithms can be equivalently derived for the complex domain, such as the gradient-based and Newton update algorithms in a very straightforward manner. We then emphasize the importance of incorporating complete statistical information into estimation through the covariance and the so-called pseudo-covariance matrices, and the need to minimize assumptions on the statistics of complex signals such as circularity. We demonstrate how Wirtinger calculus enables one to minimize assumptions on the statistics of the complex signal, and allows one to easily incorporate the complete statistical information into the estimation.
We provide a number of examples of practical significance that demonstrate how using these tools, one can easily derive algorithms and perform their analyses, and thus truly take advantage of complex domain processing. In particular, we discuss applications in widely linear filtering, training of complex-valued multilayer perceptron networks, and independent component analysis.
Group Independent Component Analysis for fMRI Data and ICA for Joint Inference of Imaging, Genetic, and ERP dataProfessor Vince Calhoun
Vince D. Calhoun VINCE CALHOUN received a bachelor’s degree in Electrical Engineering from the University of Kansas, Lawrence, Kansas, in 1991, master’s degrees in Biomedical Engineering and Information Systems from Johns Hopkins University, Baltimore, in 1993 and 1996, respectively, and the Ph.D. degree in electrical engineering from the University of Maryland Baltimore County, Baltimore, in 2002. He worked as a Senior Research Engineer at the Psychiatric Neuro-Imaging Laboratory at Johns Hopkins from 1993 until 2002. Then he took a position as the director of medical image analysis at the Olin Neuropsychiatry Research Center and an associate professor at Yale University.
Dr. Calhoun is currently Director of Image Analysis and MR Research at the Mind Research Network and is an associate professor in the Departments of ECE, neurosciences, and computer science at the University of New Mexico. He is the author of more than 100 full journal articles, over 200 technical reports, abstracts and conference proceedings. Much of his career has been spent on the development of data driven approaches for the analysis of functional magnetic resonance imaging (fMRI) data. He has won over $12 million in NSF and NIH grants on the incorporation of prior information into independent component analysis (ICA) for fMRI, data fusion of multimodal imaging and genetics data, and the identification of biomarkers for disease.
Dr. Calhoun is a senior member of the IEEE, the Organization for Human Brain Mapping, and the International Society for Magnetic Resonance in Medicine. He has participated in multiple NIH study sections. He has worked in the organization of workshops at conferences including the society of biological psychiatry (SOBP) and the international conference of independent component analysis and blind source separation (ICA). He is currently serving on the IEEE Machine Learning for Signal Processing (MLSP) Technical Committee and has previous served as the general chair of the 2005 meeting. He is a reviewer for a number of international journals and is on the editorial board of the Human Brain Mapping and Neuroimage journals and an associate editor for the IEEE Signal Processing Letters and the International Journal of Computational Intelligence and Neuroscience.
Independent component analysis (ICA) has become an increasingly utilized approach for analyzing brain imaging data. In contrast to the widely used general linear model (GLM) that requires the user to parameterize the data (e.g. the brain’s response to stimuli), ICA, by relying upon a general assumption of independence, allows the user to be agnostic regarding the exact form of the response. In addition, ICA is intrinsically a multivariate approach, and hence each component provides a grouping of brain activity into regions that share the same response pattern thus providing a natural measure of functional connectivity. There are a wide variety of ICA approaches that have been proposed, in this tutorial we focus upon two distinct methods. The first part of this tutorial reviews the use of ICA for making group inferences from fMRI data. Unlike univariate methods ICA does not naturally generalize to a method suitable for drawing inferences about groups of subjects. For example, when using the general linear model, the investigator specifies the regressors of interest, and so drawing inferences about group data comes naturally, since all individuals in the group share the same regressors. In ICA, by contrast, different individuals in the group will have different time courses, and they will be sorted differently, so it is not immediately clear how to draw inferences about group data using ICA. Despite this, several ICA multi-subject analysis approaches have been proposed. The various approaches differ in terms of how the data is organized prior to the ICA analysis, what types of output are available (e.g. single subject contributions, group averages, etc), and how the statistical inference is made. In this tutorial we first provide an introduction to ICA and ICA of fMRI, then we provide an overview of current approaches for utilizing ICA to make group inferences and also show example of how group ICA can be utilized to make inferences from fMRI data. We will provide a number of examples in which ICA has been used to discover new information about the healthy and diseased brain. In addition we will discuss several available software packages with a focus upon the group ICA approach implemented in the GIFT software.
In the second part of this tutorial, we discuss an ICA-based framework to combine or fuse multimodal data in groups of subjects using features extracted from the single-subject data. Many studies are currently collecting multiple types of imaging data from the same participants. Each imaging method reports on a limited domain and typically provides both common and unique information about the problem in question. ICA has proven particularly useful for data fusion of multiple tasks or data modalities such as single nucleotide polymorphism (SNP) data or event-related potentials. The ICA-based framework is flexible and can be used to study many kinds of data. For example, in relating SNPs and fMRI data, a genetic independent component is defined as a specific SNP association, i.e., a group of SNPs with various degrees of contribution, which partially determines a specific phenotype or endophenotype. The relationship between brain function and the genetic component can be calculated as the correlation between the columns of the fMRI and the SNP mixing matrices and adaptively maximized along with the independence among components. We will discuss multiple examples including the use of complementary data such as spatially localized functional magnetic resonance imaging (fMRI) and temporally localized event-related potential (ERP) data and also fMRI and genetic data (SNP) arrays. In summary, we hope to motivate the importance of combining multimodal brain imaging data in a unified model and also to show that an ICA-based framework provides a powerful way to identify joint relationships between multimodal data which would have been missed otherwise.