The role of anatomic information in quantifying functional neuroimaging data

When using modern neuroimaging tools, such as CT, PET, SPECT, MRI and MEG, in brain research and brain diagnostics, there is a common need for including external anatomical information into the interpretation and analysis of data. This information may be used to aid the interpretation of structures in images from low resolution imaging tools. With high resolution tools it can help to identify resolved structures. It can also facilitate the merging of data from different modalities, or from different individuals. The anatomical information is often given as regions of interests (ROIs), which may be manually created from an anatomy rich image or automatically created from a standard template collection or from an atlas data base. Automatic methods will lead to a substantial reduction in bias and in size of the systematic errors. Functional ROIs can correspondingly be derived from functional images (usually PET or SPECT). Different aspects of these processes are discussed in the report.