Show simple item record

dc.contributor.author
Ozbek, Ali
dc.contributor.supervisor
Razansky, Daniel
dc.contributor.supervisor
Bölcskei, Helmut
dc.contributor.supervisor
Kozerke, Sebastian
dc.date.accessioned
2023-04-12T14:57:03Z
dc.date.available
2023-04-12T08:00:47Z
dc.date.available
2023-04-12T14:57:03Z
dc.date.issued
2023
dc.identifier.uri
http://hdl.handle.net/20.500.11850/607410
dc.identifier.doi
10.3929/ethz-b-000607410
dc.description.abstract
Optoacoustic tomography (OAT) is gaining popularity in preclinical research and clinical diagnostics due to its unique ability to visual- ize optical contrast entirely non-invasively with high spatial resolu- tion at depths unreachable by conventional optical imaging methods. State-of-the-art OAT systems are also able to operate at fast volumetric frame rates in the 100Hz range, thus providing unparalleled capabili- ties for neuroscience and cardiovascular research. However, some bi- ological dynamics, such as propagation of electromechanical waves in the heart, neural activity or animal behavior, may occur on a faster (sub- millisecond) time scale, thus requiring ever higher temporal resolution performance. In principle, OAT can map the volumetric distribution of light absorption by capturing tiny ultrasound waves produced by a sin- gle nanosecond-duration laser pulse, thus effectively producing snap- shots of living tissues with negligible “exposure” time. Consequently, the unidirectional propagation of sound waves through the imaged vol- ume determines the ultimate limitations on the frame rate. However, the information in OAT is encoded in five independent dimensions, namely the three spatial dimensions, time, and optical wavelength, commonly resulting in vast amounts of data generated when aiming at achieving high resolution in all dimensions. Hence, in practice, the temporal resolution is constrained by the data throughput capacity of the acquisition systems. This thesis aimed to exploit the multi-dimensional sparsity of OAT data in order to overcome the data throughput bottlenecks thus pushing the OAT imaging speed beyond kilohertz volumetric (3D) frame rates, unprecedented among state-of-the-art bioimaging modalities. Unlike pulse-echo ultrasonography, one unique feature of OAT images is the lack of speckles, which greatly facilitates sparse data representations in the three spatial dimensions. Significant sparsity further exist in the temporal and spectral domains. Such multi-dimensional sparsity then facilitated development of new compressed sensing acquisition and data compression approaches, thus minimizing the data required for high resolution/speed imaging. With this aim in mind, several new methodologies were developed in this thesis. First, an acquisition system enabling spatial and temporal subsampling was developed to reduce the amount of data acquired for each laser pulse. A total varia- tion (TV) based reconstruction method was implemented to reconstruct images from subsampled data without significant quality loss. The system’s performance was evaluated using numerical simulations, and validated in vivo by ultrafast tracking of freely swimming zebrafish at 1.6 kHz volumetric frame rate. Moreover, a machine learning-based re- construction method optimised for datasets with periodic motion was developed, which enabled volumetric imaging of murine heart motion by using data from a single ultrasound detector, in analogy to single- pixel camera approaches in optics. The TV-based method was also re- fined to improve spatial and temporal fidelity by implementing a new infimal convolution of total variations (ICTV) regularization strategy. ICTV reconstruction enabled resolving cardiac mechanical wave prop- agation in murine cardiac arrhythmia models at kilohertz volumetric frame rates. Finally, a universal adaptive real-time data compression and sparse data reconstruction methodologies were developed to opti- mize the storage and accelerate the transfer and reconstruction speeds of large OAT data sets produced by ultrafast imaging systems.
en_US
dc.format
application/pdf
en_US
dc.language.iso
en
en_US
dc.publisher
ETH Zurich
en_US
dc.rights.uri
http://rightsstatements.org/page/InC-NC/1.0/
dc.title
Exploring Multi-dimensional Sparsity for Ultrafast Volumetric Optoacoustic Tomography
en_US
dc.type
Doctoral Thesis
dc.rights.license
In Copyright - Non-Commercial Use Permitted
dc.date.published
2023-04-12
ethz.size
163 p.
en_US
ethz.code.ddc
DDC - DDC::6 - Technology, medicine and applied sciences::610 - Medical sciences, medicine
en_US
ethz.code.ddc
DDC - DDC::6 - Technology, medicine and applied sciences::620 - Engineering & allied operations
en_US
ethz.code.ddc
DDC - DDC::6 - Technology, medicine and applied sciences::600 - Technology (applied sciences)
en_US
ethz.identifier.diss
28158
en_US
ethz.publication.place
Zurich
en_US
ethz.publication.status
published
en_US
ethz.leitzahl
ETH Zürich::00002 - ETH Zürich::00012 - Lehre und Forschung::00007 - Departemente::02140 - Dep. Inf.technologie und Elektrotechnik / Dep. of Inform.Technol. Electrical Eng.::02631 - Institut für Biomedizinische Technik / Institute for Biomedical Engineering::09648 - Razansky, Daniel / Razansky, Daniel
en_US
ethz.date.deposited
2023-04-12T08:00:47Z
ethz.source
FORM
ethz.eth
yes
en_US
ethz.availability
Open access
en_US
ethz.rosetta.installDate
2023-04-12T14:57:04Z
ethz.rosetta.lastUpdated
2024-02-02T21:36:17Z
ethz.rosetta.versionExported
true
ethz.COinS
ctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.atitle=Exploring%20Multi-dimensional%20Sparsity%20for%20Ultrafast%20Volumetric%20Optoacoustic%20Tomography&rft.date=2023&rft.au=Ozbek,%20Ali&rft.genre=unknown&rft.btitle=Exploring%20Multi-dimensional%20Sparsity%20for%20Ultrafast%20Volumetric%20Optoacoustic%20Tomography
 Search print copy at ETH Library

Files in this item

Thumbnail

Publication type

Show simple item record