|
Satellite Neuroinformatics Dutch challenges within a global program Wednesday, June 1, 2005 4th Dutch Endo-Neuro-Psycho Meeting 2005 Doorwerth, May 31 - June 3, The Netherlands |
|
Program |
||
|
| ||
| 10.00 - 10.15 | Opening - Wytse J. Wadman - University of Amsterdam | |
| Neuroinformatics in the Netherlands | ||
|
| ||
| Chair: Wytse Wadman | ||
| 10.15 - 11.00 | Bart M. ter Haar Romeny - Eindhoven University of Technology | |
| Biologically inspired multi-scale image analysis | ||
| 11.00 - 11.45 | Ole Jensen - Radboud University Nijmegen | |
| Measuring and modeling the human beta oscillations | ||
| 11.45 - 12.30 | Rolf Kötter - H. Heine University, Düsseldorf, Germany | |
| Current challenges in the reconstruction of brain connectivity | ||
|
| ||
| 12.30 - 13.30 | Lunch | |
|
| ||
| Chair: Jaap van Pelt | ||
| 13.30 - 14.15 |
Jos Roerdink - University of Groningen
| |
| Wavelet-based data analysis in functional MRI | ||
| 14.15 - 15.00 | Klaus Linkenkaer-Hansen - Netherlands Institute for Brain Research | |
| Time series analysis: from hearts to brains | ||
|
| ||
| 15.00 | Closure | |
|
Objectives and Content
Neuroinformatics is a new research field devoted to the development of neuroscience data and knowledge bases together with computational models and analytical tools for the sharing, integration and analysis of experimental data and the advancement of theories of nervous system structure and function. A global program in Neuroinformatics is about to start, as being initiated by the OECD Global Science Forum, for the realization of a global facility for the exchange of neuroscience data, analysis and modeling tools. This program builds on National Nodes in Neuroinformatics and an International Neuroinformatics Coordinating Facility (INCF). The present session aims at further contributing to the organization of a Dutch Neuroinformatics Community and Research Program, being essential ingredients for a Dutch National Node in Neuroinformatics.Background information at www.neuroinformatics.nl. The session will emphasize Dutch contributions in the areas of (A) tool development for the management and sharing of neuroinformatics data, (B) the development of analytical and simulation tools, and (C) the development of computational models of the nervous system and neural processes.
| Abstracts
Biologically inspired multi-scale image analysis
Medical images are nowadays produced in such qualities, and societal demands for quality are so high, that a strong demand for sophisticated image analysis techniques has emerged: 'computer aided diagnosis'. There is much to be learned from the brain. In this presentation we will study a number of examples where neurophysiological findings in the early stages of vision have inspired advanced mathematical algorithms for medical image analysis: - The models for simple cells in V1 have inspired a robust class of regularized spatio-temporal differential operators for images. - The strong feedback from V1 to LGN had lead to a wealth of adaptive ('geometry-driven') diffusion algorithms for edge preserving smoothing. - The model of the multi-scale Reichardt detector for motion triggered the design of multi-scale Lie derivatives for robust optic flow detection. they outperformed all classical methods. - The revolution due to the voltage sensitive dyes has shown short range cortical column orientation connections: this has inspired context filters, and tensor voting for Gestalt grouping processes; we developed from this a new theory for the invertible orientation wavelet transform. - Finally, the multi-scale receptive field sampling evident from the retina has lead to many projects studying the 'deep multi-scale structure' of images, to study and understand the difficult hierarchical structure of images. The lecture will address these issues in an intuitive, well-illustrated manner.
Measuring and modeling the human beta oscillations
Cortical oscillations measured in humans by EEG and MEG are highly modulated by various cognitive tasks. Since these oscillations are produced by large ensembles of neurons oscillating in synchrony, they are bound to play a strong role in neuronal computation. Studies on various in vitro rat preparations have provided strong insight into the mechanism responsible for generating spontaneous oscillations. In parallel, biophysical computational models have been develop which can account for the synchronization properties observed experimentally. We set out to explore if the insights gained from the in vitro research and computational modeling apply to human cortical beta oscillations as well. We administered benzodiazepines to healthy adults and monitored cortical oscillatory activity by means of MEG. Benzodiazepine increased the power and decreased the frequency of beta oscillations over rolandic areas. To explore the mechanisms underlying the increase in beta power with GABAergic inhibition, we simulated a conductance-based neuronal network comprising excitatory and inhibitory neurons. The model accounts for the modulation of beta oscillations with benzodiazepines, implemented as an increase in GABAergic conductivity. We conclude that models developed to account for in vitro animals results on gamma and beta oscillations can be applied to account for human beta oscillations. In future work, we hope to apply the model to explain the modulations of beta oscillations observed experimentally in cognitive paradigms.
Current challenges in the reconstruction of brain connectivity
I will provide an update on current developments in databasing the brain. Since my comprehensive summary in 2001 [1], the field has enormously grown [2], has made significant progress in many respects, and is facing new challenges. I will illustrate how we address these challenges drawing upon examples from the CoCoMac database of primate brain connectivity [3].
Wavelet-based data analysis in functional MRI
Many techniques currently exist to produce images of the human brain, such as CT, MRI or PET. These imaging techniques are not only useful to obtain images of brain structure, but also of brain function. The detection of significant changes in neurological activity is not an easy task, since in general these changes are small and distributed over the whole brain, although not in the same amount. The general approach is to obtain data under both activation and resting conditions, and assess significant changes by comparing these data. In this talk, I will discuss a number of wavelet-based techniques for data analysis and visualization of functional magnetic resonance imaging (fMRI) data. First, I will present a general wavelet-based denoising scheme for fMRI data and compare it to Gaussian smoothing, the traditional denoising method used in fMRI analysis. Next the assumption of Gaussian distributed noise in the blood oxygenation dependent (BOLD) contrast, computed from fMRI time series, is critically reviewed. The noise distribution in MR images has a Rician distribution, which deviates from a Gaussian distribution, especially for small signal-to-noise ratio. The consequences of this difference will be discussed, and a new approach to obtain symmetric, nearly-Gaussian distributed noise, is discussed. I will also present a new method to extract the haemodynamic response function (HRF) from an fMRI time series, based on Fourier-wavelet regularised deconvolution and the general linear model. Results show that using subject-specific, regional HRFs significantly improves the detection of active regions in fMRI. Finally, some visualization methods for interpretation of fMRI data are presented.
Time series analysis: from hearts to brains
Healthy hearts exhibit a remarkably complex variability in beat-to-beat intervals. The development of algorithms to quantitatively characterize these time series has been facilitated by the availability of freely downloadable heartbeat data from 'PhysioNet: the research resource for complex physiologic signals' [1]. The application of those algorithms to cardiac diseases have revealed a diagnostic potential of analyzing the temporal structure and provided novel indices for constraining models of heartbeat dynamics [2]. I will review the more recent successes in applying time series analysis to neurophysiological data and describe how this approach may improve our understanding of the complex dynamics of ongoing neuronal activity in terms of underlying physiological mechanisms and functional implications [3].
|