non-metric mds analysis  (MathWorks Inc)

Journal: bioRxiv

Article Title: Behavioral and Neural Representations en route to Intuitive Action Understanding

doi: 10.1101/2021.04.08.438996

Figure Lengend Snippet: (A) To measure intuitive judgments of action similarity, participants completed an action arrangement task, during which they watched the 60 action videos and then arranged key frames from the videos according to their similarity: frames were close together if participants thought the videos were similar, or far apart if they thought they were different. (B) Plot of the first two dimensions of a Multi-Dimensional Scaling projection, to visualize broad trends in the structure in these intuitive judgments. Actions are plotted close together in this projection if participants consistently judged them to be similar.

Article Snippet: The distance matrices measured in the multi-arrangement task were averaged across individual participants and non-metric MDS was performed over this group-averaged distance matrix in MATLAB.

Techniques:

Journal: NeuroImage

Article Title: Brain networks, dimensionality, and global signal averaging in resting-state fMRI: Hierarchical network structure results in low-dimensional spatiotemporal dynamics

doi: 10.1016/j.neuroimage.2019.116289

Figure Lengend Snippet: The unthresholded group-average voxelwise correlation matrices calculated on the volume registered data were converted to distances and submitted to metric MDS analyses with two dimensions (Set 1 in left column, Set 2 in right column). The top row shows scatterplots for all voxels in the field of view of the Yeo et al. (2011) parcellations, using black dots to represent each voxel’s column in the all-to-all distance matrices. The 2- and 4-network parcellations show results by network using color for voxels that agree in their network classification. The 7-network parcellation of Yeo et al. (2011) is shown in the bottom row, with the limbic network (white in the key) rendered as black dots for visibility on a white background. Viewing the progression from 2 to 7 networks provides a concrete rendition of the hierarchical arrangement of the networks at the voxel level, with high dot densities corresponding best visually to the top levels of the hierarchical tree (2 and 4 networks).

Article Snippet: All MDS analyses compressed the full similarity space to two dimensions for ease of viewing, employed the metric version of MDS (Matlab’s mdscale ), and minimized the squared stress goodness-of-fit criterion.

Techniques:

Journal: Scientific Reports

Article Title: Dynamic representation of time in brain states

doi: 10.1038/srep46053

Figure Lengend Snippet: Spatiotemporal dynamics and temporal perception ( a ) Similarity index of the spatiotemporal dynamics across trials (mean ± s.e.m.). Red line at the bottom represents the temporal cluster where similarity was significant. ( b ) Pairwise multivariate distance matrix between all time points. There is a strong similarity for time points closer in time, suggesting a sequence of activation states. ( c ) Multidimensional distance between time points for the period between 0.65 s and 2.2 s visualised in two dimensions using multidimensional scaling (MDS). The colour of each point represents its physical interval. The trajectory represents the path linking the sequence of activation states.

Article Snippet: Trajectory in state space was visualised using metric multidimensional scaling (MDS) as implemented in MATLAB.

Techniques: Sequencing, Activation Assay

Journal: Scientific Reports

Article Title: Dynamic representation of time in brain states

doi: 10.1038/srep46053

Figure Lengend Snippet: Distance in state space is correlated to distance in time ( a ) Multidimensional distance between the activity in the different possible intervals visualised in two dimensions using multidimensional scaling (MDS). The colour of each point represents its physical interval. ( b ) Mean distances in state space (Mahalanobis distance) as a function of temporal separation ( log 10 scale). Blue (red) markers shows pairwise multivariate distances (mean ± s.e.m.) between the 0.8 s (2.27 s) and all other intervals. The slope of the fitted linear functions indicated that the rate of change in state space as a function of time is faster for the first than for the last interval.

Article Snippet: Trajectory in state space was visualised using metric multidimensional scaling (MDS) as implemented in MATLAB.

Techniques: Activity Assay