Diosdi, A.; Toth, T.; Grexa, I.; Schrettner, B.; Hapek, N.; Kovacs, F.; Kriston, A.; Harmati, M.; Buzas, K.; Pampaloni, F.; Piccinini, F.; Horvath, P.

Multicellular 3D biological models, the so-called -oids, are the pivot key for the new generation of high-content screening (HCS) of drug analysis, cancer research, and regenerative medicine. However, the standardisation of 3D cell culture generation, handling, imaging, and data analysis remains a challenge and lacks convincing applications. In this work, we propose HCS-3DX, a next-generation system revolutionising HCS research in 3D imaging and evaluation. HCS-3DX is based on three main components: an automated Artificial Intelligence (AI)-driven micromanipulator for oid selection, an engineered HCS foil multiwell plate for optimised imaging, and an image-based software for single-cell data analysis. The developed system was validated through different experiments with 3D tumour models, including tumour-stroma co-cultures. The results prove that the resolution achievable with HCS-3DX enables us to overcome the limitations of current systems and reliably perform 3D high-content screening (HCS) at the single-cell level.