MDSS Plot Plugin Help

Author: Fabrizio Esposito

Multi-dimensional similarity scaling (MDSS) is a statistical tool to visualize the (dis)similarities between pairs of objects in which the objects are represented as points in a low dimensional space and the dissimilarities as distances between these points with the fundamental constrain that the distances between points correspond as closely as possible to the dissimilarities between the objects. In the MDSS plot plugin these objects correspond to spatially distributed brain activity patterns originating from several fMRI observations in a given anatomical space, e. g. multi-voxel patterns in the volume space (volume maps) or multi-vertex patterns in the cortical space (surface maps).

 

Because the original dissimilarities (i. e. distances in high dimensional space) are obtained from a huge number of dimensions, only a small amount of the total covariance explained by these patterns in the original space can be correctly represented in a two-dimensional configuration, therefore geometrical distortions will systematically occur in the plot. Nonetheless, in many circumstances, the resulting distribution will tend to capture the main structure of the comparative similarities and the generated plots will be a trustworthy representation of the maps' similarities at a very "global" level, a potentially useful aspect for the qualitative exploration and representation of multivariate fMRI patterns. As far as the resulting distribution of points will capture this structure, the MDSS plot display may function as an interactive and flexible visual tool for searching "intrisc" clusters in the data and for detecting outliers.

For more details about the theoretical background, and for an illustration of the practical usage, of the MDSS plugin please click on one of the following links below:

References

Friston, K.J., Frith, C.D., Fletcher, P., Liddle, P.F., Frackowiak, R.S., 1996. Functional topography: multidimensional scaling and functional connectivity in the brain. Cereb. Cortex 6, 156-164.

Kherif, F., Poline, J.B., Meriaux, S., Benali, H., Flandin, G., Brett, M., 2003. Group analysis in functional neuroimaging: selecting subjects using similarity measures. NeuroImage 20, 2197-2208.

Himberg J, Hyvarinen A, Esposito F. Validating the independent components of neuroimaging time series via clustering and visualization. Neuroimage. 2004 Jul;22(3):1214-22.

Esposito F, Scarabino T, Hyvarinen A, Himberg J, Formisano E, Comani S, Tedeschi G, Goebel R, Seifritz E, Di Salle F. Independent component analysis of fMRI group studies by self-organizing clustering. Neuroimage. 2005 Mar;25(1):193-205.