Project

Back to overview

Hierarchical X-ray imaging of the entire human brain

English title Hierarchical X-ray imaging of the entire human brain
Applicant Müller Bert
Number 185058
Funding scheme Project funding
Research institution Departement Biomedical Engineering Universität Basel
Institution of higher education University of Basel - BS
Main discipline Other disciplines of Physics
Start/End 01.04.2019 - 31.03.2023
Approved amount 855'499.00
Show all

All Disciplines (2)

Discipline
Other disciplines of Physics
Biomedical Engineering

Keywords (6)

brain anatomy; micrometer resolution; big data; hard X-ray tomography; paraffin embedding; three-dimensional registration

Lay Summary (German)

Lead
Das menschliche Gehirn wiegt 1,3 kg und enthält 86 Milliarden Zellen. Kürzlich haben Forscher mit Hilfe der Histologie einen Atlas des Gehirns aus 20 µm grossen würfelförmigen Bausteinen erstellt. Nur für Bausteine kleiner als 5 µm oder sogar 1 µm werden einzelne Zellen sichtbar. Es stellt sich die Frage, welche Ortsauflösung für die Sichtbarmachung des Hirns erreichbar ist. Dazu schlagen wir die Röntgentomographie vor, um das physische Schneiden des Gehirns zu vermeiden. Dieses Vorgehen wurde bereits 2018 für ein Mäusehirn angewandt und für die Speicherung 40 TB benötigt. Weil das menschliche Gehirn aber 3000-mal grösser ist, erwarten wir eine Datenmenge von 100 PB.
Lay summary

Wir werden zuerst ein menschliches Gehirn für die Röntgentomographie vorbereiten, dann ein Verfahren entwickeln, mit dem wir Objekte, die 100000-mal grösser als die Bildpunkte des Detektors sind, tomographisch abbilden und damit einen Atlas mit der bestmöglichen Ortsauflösung erstellen. Sachkundige finden, dass 5 µm bereits eine grosse Herausforderung darstellt. Wir wollen jedoch mindestens 1 µm erreichen, um jeden 20 µm grossen Baustein aus nahezu 10000 Elementen aufzubauen. Diese riesigen Datenmengen werden der Allgemeinheit für Forschung und Lehre frei zur Verfügung stehen.
Das Verständnis und die Behandlung der Erkrankungen des Gehirns hängt von der Verfügbarkeit detaillierter dreidimensionaler Daten ab. Deshalb wird der vorgesehene Atlas helfen, neuro-degenerative Erkrankungen, die immer häufiger in unserer alternden Gesellschaft auftreten, zu entschlüsseln. Die hierarchische Bildgebung und die Handhabung riesiger Datenmengen ist ein Paradigma für das Studium anderer Organe, hochgezüchteter technischer Produkte und einmaliger Kulturgüter.

Direct link to Lay Summary Last update: 04.04.2019

Lay Summary (English)

Lead
The human brain weighs 1.3 kg and contains 86 billion cells. Using histological sections and optical microscopy, researchers recently provided a 1 TB brain atlas with 20 µm-wide voxels. However, individual cells will remain invisible until 5 or even 1 µm-wide voxels are reached. The question arises what is the highest spatial resolution achievable for the visualization of the entire human brain? Our team aims to address this with state-of-the-art X-ray tomography that avoids serial sectioning. This approach was very recently applied to a full mouse brain, but the human brain is 3,000 times larger. The related data would increase from 40 TB to about 100 PB.
Lay summary

The goals of this project are (i) to suitably prepare the human brain for post-mortem imaging using X rays, (ii) to develop a procedure for tomographic X-ray imaging of an object that is 100,000-times larger than the pixel size, and (iii) to generate an atlas of the entire human brain with the highest possible spatial resolution. Experts in the field believe that reaching a voxel size of 5 µm is already a formidable challenge. Our team, however, will pursue 1 µm resolution and below, which would provide a voxel size nearly 10,000-times smaller than currently available. These big micro-anatomical brain data will be made freely available to the scientific community for more detailed analysis and teaching purposes.
Understanding and treatment of brain-related diseases depend on the availability of detailed three-dimensional data. Therefore, the envisioned brain atlas will assist in decoding neurodegenerative diseases that are increasingly prevalent in our ageing society. The hierarchical imaging procedure and the handling of the big imaging data will create a paradigm for investigating other human organs, high-performance engineering devices, and unique cultural heritage objects.

Direct link to Lay Summary Last update: 04.04.2019

Responsible applicant and co-applicants

Employees

Associated projects

Number Title Start Funding scheme
144535 Tomography of microvascular structures in murine brain tumors 01.01.2013 Project funding
133802 Micro- and Nanotomography 01.03.2011 R'EQUIP
150164 Multi-modal matching of two-dimensional images with three-dimensional data in the field of biomedical engineering 01.11.2013 Project funding
127297 High-resolution phase contrast micro computed tomography of soft tissues 01.10.2009 Project funding
125406 High-resolution 3D imaging of the human brain post mortem 01.04.2009 Interdisciplinary projects
147172 Micro- and nanoanatomy of human brain tissues 01.09.2013 Interdisciplinary projects

Abstract

The human brain with a weight of 1.3 kg contains around 1,000,000,000,000 cells. Current protocols for tissue imaging with cellular resolution involve optical and electron microscopies. Here, three-dimensional imaging requires serial sectioning due to the limited penetration depth. Sectioning, however, is destructive and artifact prone, with insufficient spatial resolution (perpendicular to the cuts) for visualizing cells. In the last decade, hard X-ray tomography has created virtual histology for imaging biological tissues with isotropic voxels at the micrometer scale and below. The question now arises: what is the highest achievable resolution for an atlas of the entire human brain? Currently, the best atlas is based on histological sectioning with 20 µm-wide voxels. Recent studies using synchrotron radiation-based hard X rays have imaged a full mouse brain with 0.8 µm pixel size, but the volume of a human brain is 3,000 times larger.The goal of this project is to generate an atlas of the entire human brain with 1 µm resolution. Hierarchical synchrotron radiation-based X-ray imaging will be performed at the beamline BM18 (European Synchrotron Radiation Facility - ESRF, Grenoble, France), which is under construction and will be unique with respect to beam size and transversal coherence. Here, full-field tomography of the entire human brain with voxel sizes of 20 µm and stitched local tomography acquisitions can be combined. The high-resolution imaging of the at least 12 cm-wide human brain poses several challenges, including sample preparation to withstand extreme radiation dose, acquisition protocols to keep the beamtime within limits, and data processing, which comprises slices of 100 GB and a total volume in the PB range.As the synchrotron radiation-based high-resolution X-ray measurements of the human brain can only be performed post mortem, we will make a correspondence from the ex vivo conditions back to the in vivo case in order to put this dataset into a more physiologically relevant context. Non-rigid registration of lower resolution magnetic resonance images of the human brain taken before extraction and throughout the fixation process allow for the quantification of the local deformations introduced during extraction, fixation, and embedding. In addition, the periodic high-resolution X-ray imaging of a mouse brain during the die-off process and subsequent tissue fixation and embedding will enable us to reasonably correct the brain’s microanatomy. The synchrotron radiation-based X-ray imaging of the mouse brain with less than 1 µm voxel size will be carried out at the Biomedical beamline ID17 (ESRF, Grenoble, France), which is dedicated to biomedical imaging and radiation therapy.We want to make the big data yielding the microanatomy of the entire human brain freely available to the scientific community and for teaching purposes. This dataset will provide vital advances of relevance to a systems-level understanding of the human brain, maybe even close to physiological conditions. The hierarchical imaging procedure and the handling of the big imaging data is exemplary for further applications including comparatively studying entire healthy and diseased organs, industrially relevant engineering devices, and unique cultural heritage objects.
-