overview: Using ultrasound techniques for brain imaging is potentially less harmful and more cost-effective than other current neuroimaging methods.
sauce: ETH Zurich
Both ultrasound for medical imaging and seismology for imaging inside the Earth measure the propagation of waves through matter. For example, when seismic waves encounter material differences inside the Earth (such as between different rock layers), they are reflected and refracted at interfaces. As a result, the wave speed changes.
When researchers measure these waves at the surface, they can draw conclusions about the structure of the Earth’s interior, rock composition, material properties such as density, pressure and temperature.
With the help of advanced algorithms and high-performance computers such as CSCS’ “Piz Daint”, such as Andreas Fichtner, ETH Zurich, Professor and Head of the Seismology and Wave Zoology Group at the Institute of Geophysics Researchers can use this wave data to: Characterize the three-dimensional structure of the Earth.
Due to the similarities in propagation between ultrasound and seismic waves, and the team’s know-how in the field of wave dynamics (how to use the information that waves carry and convert them into images), Professor ETH and his group also conducted research. I was. Wave propagation of medical ultrasound.
The researchers continue to work with doctors at the University Hospital of the University of Zurich to further develop these techniques.
If Marty succeeds in further developing brain meshing and imaging procedures over the next three years of his doctoral dissertation, these same methods could potentially be applied to other parts of the body, such as knees and elbows. there is.
This serves as a promising basis for developing corresponding ultrasound devices.
Patrick Marty, a PhD student in Fichtner’s group, is currently developing a method to overcome this challenge in his doctoral dissertation, with the assistance of Christian Böhm, senior scientist in the Seismology and Wave Physics group. . Scientists say the method should provide the basis for high-resolution brain imaging with ultrasound.
To simulate the propagation of waves through the brain, researchers are developing algorithms that perform many computations on a special grid known as a ‘mesh’. At its core is a software package called Salvus. Developed at ETH Zurich with the support of CSCS, Salvus models the propagation of complete wave fields (whole waves) over spatial scales from millimeters to thousands of kilometers.
ETH seismologists use this software to simulate seismic waves. For example, it is used to explore the interior of Earth and Mars, and is now used for medical imaging.
This software package uses the spectral element method (SEM). It is particularly suitable for simulating wave propagation in media with high-contrast material transitions, such as soft brain tissue and bone.
“Unlike conventional ultrasound, which only uses wave arrival times, the simulation uses the entire wave information,” Marty says.
This means that the wave shape, frequency, velocity and amplitude at each point of propagation are factored into the calculation.
Learn with a Magnetic Resonance Imaging Scanner
In their model, researchers first use an MRI of the brain as a reference. He then performs calculations on the “Piz Daint” supercomputer using different parameters until the simulated images match those of his MRI.
Using this method yields a quantitative image rather than the low information content grayscale image typical of conventional ultrasound.
By using all the information from the complete wave field, researchers can accurately determine the physical properties of the medium—the speed at which ultrasound propagates through tissue, attenuation properties, and tissue density—at all points in the medium. can be mapped to brain.
This makes it possible to finally identify the tissue type and distinguish whether it is a brain mass or a tumor tissue. This is because laboratory experiments know the density, attenuation, or velocity of sound associated with different types of tissue.
The researchers believe that this method can be used to distinguish between healthy and diseased tissue in a non-invasive and cost-effective manner. Specifically, the method can be fed into a computer integrated into an ultrasound machine specially developed for this purpose.
A computer uses the ultrasound signals recorded by the sensors to perform a series of calculations, the result of which is a 3D image of the examined brain. However, researchers stress that this still has a long way to go before it enters the clinical setting.
A particular remaining challenge is the complex geometry of the skull, such as the eye, nose, and jaw cavities, which must be accurately modeled in simulations without dramatically increasing computation time.
To solve this problem, Marty develops a method to create individual numerical meshes of arbitrary skull shapes from hexahedrons (small elements with six faces).
“Using these small deformed cubes is 100 to 1000 times faster than using tetrahedrons,” says Böhm.
“Furthermore, the project has benefited greatly from new developments in graphics cards, such as those found in Piz Daint and, in the future, Alps. They are ideal for this method. ”
So about six years ago, a research group worked with doctors to successfully develop an ultrasound method for early detection of breast cancer. The team is currently investigating ways to examine the brain with ultrasound. This method will allow researchers and doctors to monitor stroke patients and identify brain tumors, for example.
Non-invasive, cost-effective testing
Compared to computed tomography (CT) and X-rays, ultrasound has decisive advantages. This means that the procedure is mostly harmless to the body. Moreover, much more cost-effective than, for example, magnetic resonance imaging (MRI), ultrasound machines are mobile for use in remote locations.
The problem, however, is that ultrasound has so far only worked well for soft tissue.Bone strongly reflects and attenuates waves, making it difficult for ultrasound to reach through hard structures such as the skull. is very difficult.
About this Neuroimaging Research News
author: Simone Ulmer
contact: Peter Ruegg – ETH Zurich
image: Images credited to Marty, P. et al.
Original research: closed access.
Marty, P. et al. medical imaging physics
Full Waveform Ultrasound Modeling of Soft Tissue-Bone Interactions Using a Conforming Hexahedral Mesh
Full waveform modeling serves as the basis for many new inversion techniques in ultrasound CT. The ability to accurately delineate strong material interfaces, such as between soft tissue and bone, is particularly important to ensure that these numerical methods produce physically correct results.
We present a procedure for constructing digital twins of different parts of the human body using hexahedral mesh fitting. It is used in conjunction with the spectral element method to accurately model the interaction of ultrasonic wave fields at these sharp material boundaries. In silico skull and knee phantoms are used as examples.