The automation of fluorescence microscopy represents a challenge for the study of rare or transient events in biology and medicine. It relies on intelligent devices that integrate and interpret observed data and react to targeted biological events. The Roboscope (MRIC facility) is a new autonomous microscope combining interruption of the acquisition flow and event detection by deep learning. This system is distinguished by the parallelization of tasks, simultaneously performing image acquisition and data analysis to accelerate event-driven acquisition.
Nikon confocal microscopes (MicroPICell facility) enable adaptive feedback mmicroscopy, a form of smart microscopy where image analysis is performed during acquisition, and the results are used to automatically adjust microscope settings or guide the experiment in real time. By integrating user-designed advanced analysis workflows directly into the acquisition process, this approach makes it possible to detect rare or highly specific events and automatically adjust the microscope settings in real time.