Figure 5
Combined contingency table and linear regression between absolute monthly change in discrimination index and retinal layer thickness measured by OCT in 4-month-old C57BL/6J mice. Discrimination indices were quantified by a NORT at 4 months of age, 6 months of age and 9 months of age. The absolute monthly change in discrimination indices was calculated by the difference in discrimination indices divided by the time elapsed between the measurements in 9 m vs 6 m and 6 m vs 4 m of age. Based on the DSM-V, the threshold of NCD was set as an absolute monthly change in discrimination index less than one standard deviation below the mean, which was − 0.2222. Mice below this threshold were considered to have NCD, and mice above this threshold were considered to have no NCD. The threshold for the absolute change in INL thickness for the diagnostic test was interpolated using the INL regression line equation in Fig. 3 (y = 3.046x − 0.5284), which was − 1.205 pixels per month. Mice above this threshold for the absolute change in INL thickness were considered to have a positive diagnostic test outcome, and mice below this threshold had a negative test outcome. 21 of 24 mice had NCD, while 3 of 24 mice had no NCD. All mice with NCD had a negative test outcome. 18 of 21 mice without NCD had a positive test outcome. None of the mice with a negative test outcome had an absolute monthly change in discrimination index higher than zero. The contingency table was analyzed with a Fisher’s exact test for statistical significance (p = 0.0099**). Sensitivity = 0.8571, specificity = 1.0000, positive predictive value = 1.0000, negative predictive value = 0.5000. Data points represent values for one mouse at either six months vs baseline or 9 months vs baseline. INL = inner nuclear layer.
