PROJECT SUMMARY Age-related macular degeneration (AMD) is the leading cause of severe vision loss in the developed world. AMD is characterized by progressive degeneration of photoreceptors and the retinal pigment epithelium (RPE) that may be accompanied by the development of macular neovascularization (MNV). MNV gets clinically apparent when these new vessels become ‘exudative’. This exudative stage of ‘neovascular’ AMD can be treated by application of biologics that inhibit vascular endothelial growth factor. Without treatment, exudative MNV typically results in extensive fibrosis with severe central vision loss. Interestingly, there is a growing body of evidence that photoreceptor and RPE degeneration is slowed down in eyes exhibiting ‘non-exudative type 1 MNV’ that is located in the sub-RPE space. However, the proof of a direct protective effect of this specific MNV sub-type on visual function in AMD is lacking. The proposed project aims to demonstrate (1) relative preservation of retinal function along with (2) preserved structure in the immediate vicinity of type 1 MNV, whereas progressive loss of sensitivity and degeneration occurs in the surrounding tissue. This will be approached by the application of in vivo high-resolution retinal imaging combined with spatially- resolved testing of retinal sensitivity by fundus-controlled perimetry. Furthermore, given the need to minimize the burden of laborious functional testing on patients and investigators in future studies in AMD, (3) artificial intelligence will be applied in order to infer retinal function from high-resolution retinal imaging. Aim is to identify a minimum set of clinical procedures that still allows estimation of spatially-resolved retinal sensitivity and that may be implemented in future study protocols. This research project challenges the current concept of AMD. Substantial evidence for a protective potential of type 1 MNV would support the view that MNV development is actually an intrinsic ‘rescue mechanism’. This project will provide essential information to interpret current treatment approaches in AMD and would also provide a rationale for controlled MNV induction as therapeutic strategy to prevent vision loss in AMD.