Getting started with VoxBo
From VoxBoWiki
We are working on a tutorial, complete with sample data. When it's done, we'll post to the mailing list and make it available via this wiki. However, this overview will get you situated.
Once you've collected your data, there's a reasonably stereotypical series of tasks you'll want to tackle.
Contents |
Life cycle of data in VoxBo
Get your data onto the analysis computer or cluster in a format VoxBo understands
We offer relatively little help with this, but check with the list of VoxBo's Supported file formats and see what people at your imaging center are using for file formats. Chances are there's some overlap. It's usually not too hard to get your data into either Analyze or DICOM format. We also maintain a list of file format conversion tools.
Preprocessing
Preprocessing includes relatively content-independent things, like smoothing and motion correction, that are carried out before formal statistical analysis. In VoxBo, there are various scripting tools, some graphical and some text-based, that will allow you to create, edit, and run preprocessing scripts on your data.
Right now the easiest way to get started is to create a vbprep script. This is a text file that includes the list of steps you want applied to your data and the list of data files you want processed. It's easiest to explain by example, and we provide an example script broken down into two pieces with the VoxBo distribution. The files are usually in /usr/local/VoxBo/scripts, but you can also view them on this wiki:
bigprep.vbp has the processing steps, while bigprep_example.vbp has your data and parameters. Breaking it down like this makes it easy to use the same script with multiple datasets.
When you're done setting up your script, submit it by making sure both files are in your current directory and doing:
vbprep bigprep_example.vbp
Although editing text files has a lot of advantages, this is a somewhat unpolished procedure! We are working on a graphical interface and workflow model that will make this much more intuitive.
Statistical Modeling
In VoxBo, this usually means univariate analysis with the modified general linear model. VoxBo provides both text-based and graphical tools for designing a statistical model and applying it to your data.
Visualization and Diagnosis
This includes a broad range of mostly interactive procedures for examining your results, displaying maps of inferential statistics, 'surfing' voxel time series, and producing figures for publication.
Publishing
Don't forget to cite VoxBo!
