Whole Brain Mapping with X-rays

Prof. Yeukuang Hwu from Institute of Physics, Academia Sinica

@ Chin-Pao Yang Lecture Hall (Room 104), CCMS-New Physics Building

Abstract:

 

The complexity and sheer mass of the complete neural networks in an animal brain is beyond the current technology to describe and analyze. Comprehensive mapping of neural networks in the brain is therefore a formidable but very exciting challenge. Could x-ray techniques be the tool of choice to challenge the animal brain circuitry mapping? Is the overall performance adequate, however?

I will present an effective strategy based on recent advances to put synchrotron x-ray tomography into this competitive race. The approach reaches three critical objectives: (1) three-dimensional (3D) imaging with high and isotropic spatial resolution; (2) fast image taking and processing, as required for comprehensive whole-brain mapping within a reasonable time, and (3) multi-scale resolution, to zoom into specific regions of interest. The current performance: 0.3 μm lateral resolution isotropic in all 3 dimensions and image acquisition speed of 1mm^3/min in high throughput microtomography; 20 nm resolution with (50 μm)^3/min speed of nanotomography, is orders-of-magnitude faster than the competing 3D imaging techniques. We tested the strategy by mapping large populations of metal-labeled neurons and their connections in two animal models, Drosophila and mouse. Its speed notably allowed full 3D mapping of the Drosophila brain in a few days.

These positive results instigated two additional directions for further improved: an even better spatial resolution and higher probe depth, both are relevant to the high brightness synchrotron radiation and new nanofabrication facilities. An ongoing project also aims to improve the heavy metal staining efficiency and specificity. With these improvements in place and the newly initiated SYNAPSE (Synchrotron for Neuroscience – an Asia Pacific Strategic Enterprise) consortium with 6 synchrotron facilities, the combined data acquisition and processing power, we are confident that a human brain can be mapped within 3 years.

<< Back