Photoreceptor transplantation represents a promising approach to treat photoreceptor degeneration. However, improved transplantation outcome in pre-clinical animal studies relies on donor cell enrichment, which is usually achieved via fluorescence- or magnetic-activated cell sorting. Both methods require genetic modifications or the use of yet largely undefined, specific cell surface binding antibodies, limiting their use in clinical application. An alternative approach is to identify and sort cells depending on their inherent morpho-rheological properties. Using our recently introduced high-throughput real-time deformability cytometry (RT-DC) system, we showed that differences in physical and mechanical features such as size and compliance are sufficient to distinguish murine primary rods within an unlabeled retinal cell population. Here, we will employ a fluorescence-equipped RT-DC system to analyze the morpho-rheological characteristics of human retinal organoid photoreceptors generated from cone- and rod-specific fluorescent reporter iPSC-lines. Finding criteria that maximize the discrimination of cones or rods will be supported by machine and deep learning. Using these criteria, unlabeled photoreceptors will be sorted and transplanted into pre-clinical mouse models of retinal degeneration, allowing examination of their therapeutic potential. Together, this project represents an innovative and interdisciplinary approach to circumvent the modification of human target cells for separation and transplantation.