Computational 3D and 4D optical imaging at high spatiotemporal throughput
Abstract:
Conventional optical imaging systems have difficulties scaling to high spatiotemporal throughput, rendering it challenging or impossible to study complex and highly dynamic biological systems. In particular, due to physical limitations of hardware-only systems, it is difficult to develop imaging systems that can capture multidimensional information about dynamic samples at high resolution, wide fields of view, and high frame/volume rates. In this talk, I present several high-throughput computational optical imaging systems from my research that take advantage of parallelized designs and large-scale multidimensional image reconstruction algorithms to achieve spatiotemporal throughputs of several billion pixels or voxels per second. I show that these new imaging capabilities enable high-resolution, high-speed observation of the behaviors of multiple freely moving organisms in parallel, in particular ants, zebrafish, fruit flies, and C. elegans. As I start my lab at the University of Michigan, I aim to continue pushing the boundaries of computational optical imaging to achieve orders of magnitude higher spatiotemporal throughputs to accelerate scientific discovery in biology and medicine.