In charge: Anders M Fjell

We have a laboratory dedicated to analysis of neuroimaging data. At the moment, the lab consists of 10 workstations and 30 processing units, most of which runs Linux (Fedora Core 4). With the exception of use of NeuroScan software for some ERP-preprocessing and a minor part of the workstations running windows, the programs used are freely available and mostly open-source.

During the last couple of years, we have invested a lot of time, efforts, and money to develop a solid infrastructure in the lab. This includes establishment of a research server with a capacity of 12 terrabytes (thanks to Egil Jørgensen’s stamina), also connected to the ERP labs and the MR scanners at Rikshospitalet (http://www.rikshospitalet.no/). Further, the lab is connected to the University of Oslo’s cluster of processors, Titan (http://login3.titan.uio.no/ganglia/), which by Oct 2007 consisted of 448 dual-cpu-dual-core (4 cores) SUN nodes and 180 dual-cpu Dell nodes (thanks to Jostein K Sundet and Henrik Ormåsen). This makes it possible to run through enormous amounts of data processing in just a day or two, processes that would take several weeks if only the local lab processors were used.

Processing of most types of neuroimaging data can be undertaken in the lab, and the following software solutions are used:
Structural MRI: FreeSurfer (http://www.surfer.nmr.mgh.harvard.edu/) is used to analyse structural MR images. Through a close collaboration with Bruce Fischl, Doug Greve, and David Salat at A. Athinoula Center for Biomedical Imaging, Harvard Medical School, and Anders M Dale, University of California, San Diego, we have been able to use this highly powerfull software package since 2002.
Diffusion Tensor Imaging (DTI): DTI data is analysed with the diffusion toolboxes within the FreeSurfer environment, and with Tract Based Spatial Statistics (www.fmrib.ox.ac.uk/fsl/tbss/index.html) developed by The Oxford Centre for Functional Magnetic Resonance Imaging of the Brain (FMRIB).
Electroencephalography (EEG)/ Event-related potentials (ERP): Electrophysiological data is analysed with a number of different softwares. NeuroScan (http://www.neuroscan.com/) software is often used for preprocessing of data, in addition to custom made software. Through a collaboration with Scott Makeig and Arnoldo Delorme at Schwartz Center for Computational Neuroscience, we also use EEG-lab (www.sccn.ucsd.edu/eeglab) for analyses of EEG data. For source localization of high-density ERP data, dynamic Statistical Parametric Mapping (dSPM) developed by Dale, Fischl, Halgren et al. (2002, Neuron) is used, though a collaboration with among others Don Hagler at Dale and Halgren’s multi-modal neuroimaging lab at UCSD.
Functional MRI: For analyses of fMRI data, SPM (http://www.fil.ion.ucl.ac.uk/spm/), developed by the Wellcome Department of Imaging Neuroscience, and FSL (http://www.fmrib.ox.ac.uk/fsl/) , developed by the FMRIB group at Oxford, are used.