Next: 1.1.1 Models Available in Up: 1. Introduction Previous: 1. Introduction
Lyngby is a Matlab toolbox for the analysis of functional magnetic resonance imaging (fMRI) time series. The main purpose of the toolbox is to model four-dimensional fMRI data (i.e. 3D spatial volume over time) and to derive parameter sets from them that will allow easy interpretation and identification. The toolbox was primarily written for the analysis of experimental data obtained from controlled multicentre trials - The Human Brain Project ``Spatial and Temporal Patterns in Functional Neuroimaging'', carried out by the International Consortium for Neuroimaging. All of the methods have low-level modelling functions and a graphical user interface (GUI) interface for easy access to the data and modelling results. It is important to realize that no single model can grasp all the features of the data. Each of the models have their own contributions, and the assumptions underlying the models are very different in nature.
The Lyngby toolbox can import data in various different formats, and it is also able to cope with non-supported formats with very little user-intervention. In addition to the large set of data modelling strategies, there is also a choice of pre-processing steps and routines, and the ability to perform post-processing on the modelling results.
The toolbox was developed on Linux and SGI platforms and should work on all platforms with Matlab version 5.2. Lyngby was previously supported with Matlab 4.2 and some of the functions will still work with this version. There are still some teething problems with running the toolbox under MS Windows however, due to some differences in the way Matlab operates between the Unix and Windows flavours. We are working to overcome these problems, but as they are due to subtle, undocumented Matlab 'inconsistencies', it is something of a trail-and-error process. For the time-being, we encourage people to use the toolbox in Matlab under Linux, where the majority of our development is now carried out. Any advice regarding the development of the toolbox in Matlab for MS Windows is, of course, welcomed.
In addition to this manual, there are also a few other pieces of documentation to aid use of the toolbox. A ``Getting Started'' guide has been written to show the first-time user how to download, install and run the toolbox as quickly and simply as possible. In addition, an ``Example Analysis'' guide demonstrates how Lyngby can be used to analyse one of the sample datasets, extract some meaningful results, and then interpret them. Both these documents are included in this manual as appendices.