Skip to main content
. Author manuscript; available in PMC: 2012 Sep 1.
Published in final edited form as: Ann N Y Acad Sci. 2011 Sep;1233(1):320–326. doi: 10.1111/j.1749-6632.2011.06118.x

Figure 4.

Figure 4

Panel A shows a scatter diagram of the change in the SD of the velocity about each axis of rotation (symbols) for each drug (colors) in all OPT patients. Only gabapentin and memantine showed a consistent reduction in gain (i.e., the slope between the pre- and post-medication velocity of SDs). Regression lines: baclofen = 8.63 + 0.28 * control, R2 = 0.11 (blue); memantine = 3.77 + 0.30 * drug, R2 = 0.47 (magenta); gabapentin = 3.99 + 0.26 * control, R2 = 0.24 (green). In panel B, the values of the frequency range containing 66% of the power around the dominant frequency (i.e., bandwidth) during treatment with each drug is plotted on the y-axis, while the values prior to treatment are plotted on the x-axis. The dashed line is an equality line. Each data point represents one axis of rotation from one eye in one patient. Right and left eyes are lumped together. Magenta symbols are data after memantine, green symbols are gabapentin, and blue symbols are baclofen. Circles are horizontal component, squares are vertical component, and triangles are torsional component of the data. A broader bandwidth corresponds to more cycle-to-cycle irregularity of the frequency in the given oscillation train. Regression lines: baclofen = −0.32 + 1.05 * control, R2 = 0.81 (blue); memantine = 0.40 + 0.83 * drug, R2 = 0.71 (magenta); gabapentin = −0.24 + 1.06 * control, R2 = 0.89 (green). In panel C, box-and-whisker plots summarize the ratio of bandwidths after and before gabapentin and memantine. The horizontal lines in the center of the notch represent the median ratio, the notches represent 95% confidence intervals, the length of the box represents the inter-quartile distance, and the whiskers represent data range. Red plus symbols represent outliers. Abbreviations as in Fig. 3.