Fourier Ptychography Phenotypic Screening

Ziel

The primary objective of this project is to overcome a critical bottleneck in drug discovery by developing an ultra-fast microscopy platform capable of imaging an entire well plate at sub-micron resolution in under one minute. This will accelerate AI-based phenotypic screening by over 1,000-fold, enabling the analysis of vast chemical libraries to find new medicines for major health threats like antimicrobial resistance (AMR).

This goal will be achieved by innovating Fourier Ptychographic Microscopy (FPM), a computational technique that provides both high resolution and a wide field of view. The project integrates three approaches to optimize FPM for well-plate screening: 1) defocus-based imaging to achieve super-resolution beyond the camera's physical pixel size limits; 2) a geometry-aware algorithm to computationally correct for optical distortions introduced by the well plate's meniscus and walls; and 3) a compressed-sensing strategy that exploits the sparse nature of biological samples to reconstruct high-quality images from significantly fewer captures, enabling high speed.

The project is highly relevant as it directly addresses the urgent need to accelerate preclinical research and supports EU health priorities, particularly the fight against AMR. By developing a breakthrough imaging technology and disseminating it via open-science practices, this work will strengthen European leadership in biophotonics and AI-driven life sciences and has strong potential for impact in other fields like digital pathology and industrial inspection.