Image analysis for eyecare

We are developing methods for the automated analysis of image data generated in ophthalmology. The expertise we have acquired in medical image processing enables us to offer ophthalmologists and manufacturers of ophthalmological devices valuable assistance in evaluating the data obtained. In this way, the otherwise laborious manual processes can be automated (for example with the help of deep learning methods), and physical features can be extracted for use by the medical experts in arriving at their diagnoses.

Our experience in this area includes many different imaging techniques such as optical coherence tomography (OCT), sonography (ultrasound), fluorescence angiography and retinal images employing adaptive optics (AO). In AO images, for example, the methods we have developed can be used to automatically segment the vessels and thus determine their diameter. Furthermore, individual cones can be detected, allowing cone density on the retina to be determined – a vital factor in the detection of many diseases.

OCT image of the retina
Calculated density distribution of the cones
Colour-coded representation of the diameters of the blood vessels


3D printing is set to replace the purely manual production of prosthetic eyes, heralding a revolution in this specialist field. This disruptive, ground-breaking 3D-printing technology has been developed by Fraunhofer IGD. It significantly accelerates the manufacture of prostheses, and entails faster, better and less invasive procedures. The first patients received their fully 3D-printed eyeballs in late 2021.

Fraunhofer IGD developed Cuttlefish:Eye in close collaboration with Ocupeye Ltd in Kenilworth and Moorfields Eye Hospital in London. Using a 3D scan of the eye socket and a color-calibrated photo of the healthy eye, this powerful software generates a highly accurate virtual model suitable for the 3D printing process.

© Moorfields Eye Hospital NHS Foundation Trust