Abstract
For some experimental approaches in brain imaging, the existing normalization techniques are not always sufficient. This may be the case if the anatomical shape of the region of interest varies substantially across subjects, or if one needs to compare the left and right hemisphere in the same subject. Here we propose a new standard representation, building upon existing normalization methods: Cgrid (Cartesian geometric representation with isometric dimensions). Cgrid is based on imposing a Cartesian grid over a cortical region of interest that is bounded by anatomical (atlas-based) landmarks. We applied this new representation to the sensorimotor cortex and we evaluated its performance by studying the similarity of activation patterns for hand, foot and tongue movements between subjects, and similarity between hemispheres within subjects. The Cgrid similarities were benchmarked against the similarities of activation patterns when transformed into standard MNI space using SPM, and to similarities from FreeSurfer’s surface-based normalization. For both between-subject and between-hemisphere comparisons, similarity scores in Cgrid were high, similar to those from FreeSurfer normalization and higher than similarity scores from SPM’s MNI normalization. This indicates that Cgrid allows for a straightforward way of representing and comparing sensorimotor activity patterns across subjects and between hemispheres of the same subjects.
Highlights
In functional brain imaging, spatial normalization is often applied, where scans are transformed into a common space, so that the same coordinates in different subjects correspond to the homologous anatomical location in the brain
We focused on four regions of interest (ROIs): left M1, left S1, right M1, and right S1
21 ± 10 tiles (1.8% of all tiles) did not contain any vertices that were labelled as being part of the sensorimotor cortex; these tiles were mostly located at the edges of the Cgrid and were excluded from the correlation analyses
Summary
In functional brain imaging (functional MRI; fMRI), spatial normalization is often applied, where scans are transformed into a common space, so that the same coordinates in different subjects correspond to the homologous anatomical location in the brain. This makes statistics at a group level possible, allowing for the comparison of brain activity patterns between groups of subjects, for example patients and healthy controls. To join multiple brain images together for comparison of brain activation between groups (for example patients versus controls) or determining common areas of activation (mapping), several options are available and widely used. Activity can be mapped on an inflated brain, where sulci are projected to a spherical surface or a flattened cortex map (Fischl et al 1999), both of which allow for subsequent normalization (Qiu and Miller 2007; Van Essen et al 2001)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
