Related Experiment Videos
Estimation of projection errors in human ocular fundus imaging.
1Lehrstuhl für Messtechnik, RWTH, Aachen, Germany.
Summary
Fundus image analysis is distorted by projection errors, especially at the image edge. A spherical eye model is sufficient for correcting these distortions within a -5 to +5 diopter range.
Area of Science:
- Ophthalmology
- Optical Engineering
- Medical Imaging
Background:
- Photogrammetric analysis of human ocular fundus images is susceptible to various errors.
- Camera and eye optics aberrations are known error sources.
- Projection distortion from the spherical fundus shape onto a planar device is often overlooked.
Purpose of the Study:
- To quantify projection errors in ocular fundus imaging.
- To investigate the influence of axial ametropia (myopia and hyperopia) on projection errors.
- To compare projection errors between spherical and ellipsoidal eye models.
Main Methods:
- Geometrical analysis of a reduced model eye.
- Imaging the model eye using a pinhole camera.
- Quantification of projection errors at different degrees of the fundus and for varying ametropia.
Main Results:
- A projection error of 24% was found near the edge of a 50-degree fundus image.
- Projection errors in hyperopia were similar to emmetropia.
- Projection errors decreased in myopia.
- Spherical and ellipsoidal eye models yielded comparable results for ametropia within -5 to +5 diopters.
Conclusions:
- Projection distortion is a significant factor in fundus image analysis.
- A simple spherical eye model adequately corrects projection distortions for common ranges of ametropia.
- Further investigation may be needed for higher diopter ranges or more complex eye shapes.