Craig R. Forest, PhD

Georgia Institute of Technology, G. W. Woodruff School of Mechanical Engineering

Committee:

R. Clay Reid, MD

Allen Institute for Brain Science

Machelle T. Pardue, PhD

Georgia Institute of Technology, Coulter Department of Biomedical Engineering

Peter J. Yunker, PhD

Georgia Institute of Technology, School of Physics

Todd Sulchek, PhD

Georgia Institute of Technology, G. W. Woodruff School of Mechanical Engineering

Batch processing of brain tissue sections for millimeter-scale serial section transmission electron microscopy connectomics

The field of connectomics has emerged a promising approach for exploring the nature of neural circuits. A millimeter-scale connectome—a neuron-to-neuron wiring diagram of a neural circuit—potentially contains significant information regarding information processing and memory. The field is held back, however, by the difficulty in consistently and rapidly collecting neuroanatomical datasets with serial section transmission electron microscopy (ssTEM). In the cerebral cortex, for instance, a local circuit is contained in a cubic millimeter, but single sections—obtained by cutting brain samples with a diamond knife—must be “ultrathin” (< 40 nanometers), thus requiring 25,000 consecutive sections to be processed. Currently, the processing of ultrathin sections remains an unsolved problem that is necessary for the advancement of ssTEM connectomics. The goals of this proposal are: (1) design, model, and test a novel device that uses hydrodynamic forces and curvature-induced capillary interactions for the transport and trapping of ultrathin sections, (2) design, implement, and characterize batch processing of single sections to enable reliable processing of thousands of serial sections, and (3) design, test, and characterize automated batched section processing, enabling high-throughput and reliable section processing. In total, these aims comprise a novel platform for section processing for millimeter-scale ssTEM connectomics studies.