Periodic Reporting for period 2 - BRILLIANCE (Bright, coherent and focused light to resolve neural circuits)
Período documentado: 2021-09-01 hasta 2023-02-28
The structure of neuronal circuits drives their function and our ability to investigate and understand the mechanisms governing their development, organization, adaptation and dysfunction has remained very limited. The access to these dense and intricate structures is conditioned by technology that can simultaneously offer nanometre scale spatial resolution and coverage of milimetre sized tissue volumes. Electron microscopy has remained the only technology capable of offering the necessary spatial resolution to resolve all neurites and synapses, but the perspective of imaging a full mammalian brain in a reasonable time frame remains out of reach with existing setups. Generating multiple maps of full neuronal circuits is essential for advancing our understanding of brain function, for finding ways to prevent and treat neurological diseases or to translate the extraordinary efficiency of the brain into novel designs of computer architectures. Our goal is to develop scalable coherent X-ray microscopy capable to image rapidly neuronal tissues with synaptic resolution.
With synchrotron X-ray holographic nanotomography we could obtain dense neuronal images of heavy metal stained mouse cortex and Drosophila tissue with sufficient quality to reconstruct neuronal architectures. In Drosophila we could reconstruct motor and sensory neurons from the ventral nerve cord and the leg. Thanks to machine learning based automated segmentation, reconstruction of neuronal architectures from these X-ray microscopy images is greatly facilitated. These first results are published in Nature Neuroscience (PMC8354006).
We developed and implemented a large sensor, high resolution indirect X-ray imaging detector to accelerate data collection in holographic nano-tomography.
We obtained evidence that 3D X-ray microscopy is capable of resolving synapses in heavy metal stained, resin embedded brain tissue samples.
With deep learning based self-supervised image restoration we could improve the spatial resolution and accelerate data acquisition in coherent X-ray microscopy.
We generated unprecedented data sets of circuit modules in mouse brain and of multiple full brains in a small model organism to investigate variability between healthy individuals and the architectural changes associated to neurological diseases.
We developed and implemented a large sensor, high resolution indirect X-ray imaging detector to accelerate data collection in holographic nano-tomography.
We obtained evidence that 3D X-ray microscopy is capable of resolving synapses in heavy metal stained, resin embedded brain tissue samples.
With deep learning based self-supervised image restoration we could improve the spatial resolution and accelerate data acquisition in coherent X-ray microscopy.
We generated unprecedented data sets of circuit modules in mouse brain and of multiple full brains in a small model organism to investigate variability between healthy individuals and the architectural changes associated to neurological diseases.
We could improve resolution of coherent X-ray microscopy to resolve the first synapses in mouse brain tissue and we could accelerate data acquisition to enable imaging of meaningful tissue volumes.
We generated unique data that enables addressing new scientific questions both related to fundamental understanding of brain function and to finding ways to tackle neurological diseases.
We generated unique data that enables addressing new scientific questions both related to fundamental understanding of brain function and to finding ways to tackle neurological diseases.