Effect of image registration on longitudinal analysis of retinal nerve fiber layer thickness of non-human primates using Optical Coherence Tomography (OCT)

Table 5 Comparison of the magnitude of residuals of the linear mixed effect model before and after registration

RNFL feature	Linear mixed effects model for residual
RNFL feature	Intercept (μm)	Slope Of *reg* _i,t	p value for slope
All rings	2.5198	0.3877	0.4075
Temporal	5.3737	−1.8789	0.0425*
Inferior	4.5483	0.0420	0.9595
Nasal	5.1851	−1.8431	0.0259*
Superior	5.6397	−1.4297	0.0773
Clock hour 01	5.3477	−1.3711	0.0944
Clock hour 02	6.0659	−1.9699	0.0359*
Clock hour 03	4.7984	−1.3750	0.0797
Clock hour 04	9.3879	−6.3046	<0.0001*
Clock hour 05	8.7581	−3.5554	0.0018*
Clock hour 06	6.5226	0.6532	0.5933
Clock hour 07	8.0791	−2.8944	0.0325*
Clock hour 08	6.5612	−2.2646	0.0488*
Clock hour 09	7.6976	−4.3798	0.0024*
Clock hour 10	8.7744	−4.9544	0.0003*
Clock hour 11	7.6400	−2.6737	0.0229*
Clock hour 12	7.1031	−2.6395	0.0060*

Coefficients and the p values of the coefficients for comparing residuals of linear mixed effects models of RNFL features vs. days before registration (reg _i,t = 0) and after registration by LPCC_MSBV algorithm (reg _i,t = 1). Most of the slopes are significantly negative (p < 0.05), which mean that the magnitude of the residuals decreased after registration.
*Slopes were negative and p values were smaller than 0.05.

ISSN: 2326-0254