Purpose:
To build a mice eye model for further retinal imaging study.
Methods:
A 9–week old mice was used for ocular imaging by a 9.4–T MRI machine (Bruker AVANCE 400, Germany). The pulse sequence was set as follow: Repetition time=5000 msec, Echo time=56 msec, Average = 240, FOV=1.5 cm, Slice thickness=0.2mm. A total of 18 slices of images including the eye and the surrounding tissues were obtained with resolution of 58*58*200 um. Technique of binary threshold to distinguish boundaries between tissues was used in the post image process. The data was then smoothed and fitted by equations for 3D qudratic surfaces which can represent for cornea, lens anterior and posterior surfaces, and retina. Criteria of minimal least square error was applied for the fittings. The curvature of a surface was estimated by selecting any five points that surround the apex of the surface and fitted by a spherical equation. The refractive power of each surface could then be calculated (the refractive index of each tissue was assumed to be similar to the humans at this moment). The thickness of cornea was estimated as 50 um according to our immunohistochemical study for the calculation of corneal refractive power.
Results:
The geometry of the mice eye is given in the Table shown below
Thus the refractive powers are: anterior surface of cornea = 370 D, posterior surface of cornea = –39 D, anterior lens surface = 120 D, posterior lens surface = 96 D. Therefore the total refractive power of the mice eye is approximately 528 D.
Conclusions:
This MRI technique can not only avoid many time consuming processes while compared to the traditional anatomical methods, which are also skill dependent, it can also provide 3D images with fairly quality of live small animals for structural measurements. This study provides a useful mice eye model that could be applied to many vision related research areas.
Keywords: imaging/image analysis: non-clinical • refraction