SFB Tomography across the scales: Adaptive optics in ophthalmic imaging
by Günter Auzinger, JKU Linz
Ophthalmoscopes are devices for non-invasive diagnostic investgation of the human eye. Especially the retina including cones and rods for visual perception as well its blood vessels and nerves are of interest for early diagnosis of various deseases. The retina is scanned by a weak laser through the eye’s pupil. Interferometric measurements of the back-scattered light allows for resolving in-depth sctructures along the laser beam, thereby a three-dimensional analysis is possible by changing the angle of the beam. During passage through the eye’s structures anterior to the retina (mainly the cornea, the lens and the vitreous body), the propagation of the laser light is perturbed due to irregularities of the refractive index, as well as movements of the eye and the lids, causing a loss of information by blurring in the scientific imaging process. For the correction of these perturbations, an adaptive optics system is applied. It consists of wave-fropnt sensors that measure the phase delays of the distorted light, deformable mirrors for compensation of these delays and a unit for data processing and control of the mirrors.
In this SFB we try to exploit synergies between ophthalmoscopy and astronomical imaging with extremely large telescopes, since adaptive optics is also utilized in these telescopes and our team has expertise in this field from cooperatioon with ESO and other institutions contributing to the ELT. Our first goal is to modify one of our control algorithms established in telescopic AO control and install it in ophthalmoscopic devices. This algorithm (CuReD) has the advantage to run very fast and stable. The critical challenge in this project is the fact that the eye is rapidly moving in a non-predictable way, causing a permanent change in the geometry of the region, where the sensor gets enough light for usable measurements. This phenomenon does not occur in astronomical imaging. Another area of research is the question, if it could pay off to replace the established sensor type (Shack Hartmann) by the newer pyramid sensor: This might have advantages due to the higher sensitivity of the pyramid sensor, since the laser for scanning the retina has to be very weak in order to avoid a demaging of the eye.
The main goal is to increase the quality of the resulting 3D-scans with respect to e.g. spatial resolution, avoidance of artefacts and acceleration of the whole examination process.