Removing barriers to advanced imaging and machine learning-based analysis

While imaging and image analysis technologies have undergone dramatic technological leaps in the last decade, progress has been uneven at the global scale(1).

Although some groups are working to bring high-value-but-lower-cost microscope hardware(2) and software(3) to end users, much work remains.

Brightfield microscopy remains popular in both high- and low-resource settings, and recent work in hardware4 and software(5,6) now extends the amount of information extractable.

In this proposal, we will leverage our groups' historical strengths in the creation of new imaging methods(7,8) and machine learning-based phenotypic analysis(9) to create hardware and software based on quantitative phase imaging (ptychography) and brightfield microscopy, focusing on time-lapse imaging of live cells.

In close partnership with sites in LMICs, we will prioritise methods that increase the amount of information researchers can extract from images while decreasing the infrastructure required to extract that information.

We will initiate two ptychography hubs in South America and Africa, to facilitate equipment access and enable co- development of a low-cost imaging-phenotyping combination for roll-out across our network.

The improvement and distribution of brightfield microscopy hardware and biologically relevant deep learning algorithms will accelerate research at sites around the globe.