Data Formats

This section introduces the different formats used for fMRI datasets and how to convert them into each other. Normally, MRI image data are stored in a binary data file as either 8- or 16-bit integers. Besides the raw image data, there is usually a metadata along with to provide the descriptive information about the subject, type of image, imaging parameters as well as image dimensions. In the history of neuroimaging there have been several different image formats playing important roles. In the following sections, three major kinds of formats are going to be discussed in detail.

• DICOM
• NIfTI
• Afni
• GIfTI

There also exists a format for organizing and describing outputs and datasets of neuroimaging experiments called the Brain Imaging Data Structure (BIDS) format^[1]. The BIDS standard uses file formats compatible with existing software, unifies the majority of practices already common in the field, and captures the metadata necessary for most common data processing operations.

Certain preprocessing workflows, such as fMRI prep, will require the data to be in BIDS format.

Preprocessing

This is a collection of possible preprocessing steps. Which steps you include and also their order depends on your research question and to an extent also on the software you use

• Data Quality
• Slice Timing
• Realignment
• Skull Stripping
• Tissue Segmentation
• Coregistration and Normalization
• Field map / bias field correction
• Physiological noise regression
• Temporal Filtering
• Spatial smoothing
• Surface extraction