Low_resolution_Thesis_CDD_221009_public - Visual Optics and ...

CHAPTER 7
We compared the changes in patients that underwent refractive surgery (Fig. 7.4
(A-F)), with normal control eyes, measured following a similar time-course (Fig. 7.4
(G-L)). In some cases, we observed statistically significant changes in control eyes
between measurements separated one day or one week, although no treatment had
been applied between the measurements. Very interestingly, left and right eyes of the
same subject (in all patients and all but one control subject) follow the same trends
with time.
7.4.4. Difference between vertical and horizontal radii and asphericities
The results reported above correspond to radii and asphericities obtained from a fit of
the topographies to rotationally symmetric ellipsoids. We have also fitted the
topographies to more general non-rotationally symmetric ellipsoids, and analyzed the
vertical and horizontal radii and asphericities separately. We did not find statistically
significant meridional differences in asphericity (Fig. 7.5 (C, D)). However, the
average radius of curvature shows statistically significant meridional differences:
while the horizontal radius does not change at any time point (Fig. 7.5 (A)), the
vertical radius of curvature suffers a statistically significant change one day after
surgery (Fig. 7.5 (B), R y = 47 m, p = 0.004). This meridional difference is
equivalent to a posterior astigmatism of 0.9 D (and an effective astigmatism of 0.037
D), and is responsible for the near-significant change of the average radius of
curvature reported before (Fig. 7.3 (A)). This meridional change cannot be attributed
to a different variability in the vertical direction for this instrument, as seen from the
validation experiments. No meridional differences were found in control subjects.
Fig. 7.5: Changes of average horizontal and vertical radii of curvature (A, B) and
asphericity (C, D). Gray dots: Patients measured the three times. Black squares: All
patients. Error bars indicate the 98.3% confidence interval. Number of eyes per point
is the same as in Fig. 7.3.
Figure 7.6 shows histograms of changes between pre-operative and all post-operative
measurements for all control eyes (A, B) and all surgical eyes (C, D). Changes in
control subjects follow normal distributions (according to Jarque-Bera normality test)
178