CelFDrive: Artificial Intelligence assisted microscopy for automated detection of rare events
Brooks, S.; Toral-Perez, S.; Corcoran, D. S.; Kilborn, K.; Bodensteiner, B.; Baumann, H.; Burroughs, N. J.; McAinsh, A. D.; Bretschneider, T.
Show abstract
11.1 SummaryCelFDrive automates high-resolution 3D imaging cells of interest across a variety of fluorescence microscopes, integrating deep learning cell classification from auxiliary low resolution widefield images. CelFDrive enables efficient detection of rare events in large cell populations, such as the onset of cell division, and subsequent rapid switching to 3D imaging modes, increasing the speed for finding cells of interest by an order of magnitude. 1.2 Availability and ImplementationCelFDrive is available freely for academic purposes at the CelFDrive GitHub repository. and can be installed on Windows, macOS or Linux-based machines with relevant conda environments [1]. To interact with microscopy hardware requires additional software; we use SlideBook software from Intelligent Imaging Innovations (3i), but CelFDrive can be deployed with any microscope control software that can interact with a Python environment. Graphical Processing Units (GPUs) are recommended to increase the speed of application but are not required. On 3i systems the software can be deployed with a range of microscopes including their Lattice LightSheet microscope (LLSM) and spinning disk confocal (SDC). 1.3 Contacts.brooks.2@warwick.ac.uk
Matching journals
The top 6 journals account for 50% of the predicted probability mass.