Group Analysis ADplot

The evaluation of differences among groups can be enhanced from the implementation of Multi-Subject ADplot.

The Group Analysis ADplot, is actaully a Muti ADplot. The difference is that for each gait graph instead of plotting a single subject indexes, the indexes of a whole group are plotted.

This version of the Multi ADplot enables the visualization of the behavior of the graphs of the group studied. Their behavior is described by the arrangement of the scattered values in the Group ADplot.

An example of Group ADplots can be viewed in Figure 1. Three Group ADplots are shown. Each one contains the sagittal plane knee graph data from three different Cerebral Palsy groups of patients. Qualitative inferences may be drawn directly by observing the scattering of the values of the groups.

Figure 1. Sagittal plane knee ADplots for a group of Hemiplegics, a group of Diplegics and a group of Quadriplegics.

In this version of the ADplot point labeling is not necessary since all points plotted are referring to the same gait graph.

In the ADplot group graphs, the minimum, maximum values of deviation and asymmetry as well as their range of values are identifiable for each CP group.

The concentration of the points in each graph can give as valuable information.

In the above example, Hemiplegic knee graphs are documented to scatter closer to normal area (gray box) than the other two groups. Most of the hemiplegic knee graphs are withing the range of 2 SD.

The trend of Diplegics and Quadriplegics to exhibit more Deviation than Asymmetry is obvious since most of their values are located well above the diagonal line. Diplegic knee graphs though are less asymmetric than Qudriplegics since a large proportion of the scattered points are bellow the value 1 SD.

Hemiplegics knee graphs also exhibit a trend of equal increase in asymmetry and deviation.

These are useful information when analyzing group differences and characteristics. This information is very comprehensive. Even non experts can spot the characteristics of the groups if they are aware of the ADplot structure.