PubMed Health⌕ Search
Home/Archive/PubMed

PubMed · 6683017

A geometric approach to the filtered backpropagation algorithm.

M L Oristaglio
View original source ↗Export metadata ↓

Abstract

We show how the filtered backpropagation algorithm [1] for parallel-beam diffraction tomography can be implemented geometrically by weighted straight-line projections. The resulting algorithm is formally equivalent to the original approach using Fourier transforms [2], but reorganizes the calculations to separately backpropagate each plane-wave component of the filtered data. In principle, at least, the projection method also suggests the design of an ultrasound scanner that records the wavefield with line detectors.

Explore related subjects

Fourier AnalysisMathematicsTomographyUltrasonicsComparative StudyJournal Article

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

M L Oristaglio. 1983. A geometric approach to the filtered backpropagation algorithm.. https://doi.org/10.1177/016173468300500104

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related citations

PubMed1998

"Hidden" sequence periodicities and protein architecture.

I demonstrate the existence in protein domain sequences of sets of statistically significant periodic signals, characteristic of the architectures of those domains. It is shown that, although the frequencies of the signals characteristic of a particular architecture have definite wave numbers, they can occur in any physical property. The characteristic signals define sequence units, which may correspond to specific structural features. Study of signals in different architectures suggests that symmetric structures are characterized by sets of sequence signals, which are members of harmonic series and which therefore impose commensurate periodicities on the structure. Asymmetric structures are characterized by mutually incommensurate sets of sequence signals. It is pointed out that, although these periodic sequence units can be very short, their existence is a property of the entire sequence.

Fourier Analysis↗
PubMed1998

Formalistic description of multislice calculation method.

This paper presents a formalistic description of the multislice method. The effects of the number of beams, the number of Fourier components for the projected potential used for the phase grating, the number of the iterations and the slice thickness on multislice calculations are discussed qualitatively. The phenomenon of "leaking" of the wave function to higher order reflections in reciprocal space in a multislice calculation is also discussed. The relationship between the number of Fourier components of the projected potential and the number of beams used for the wave function in reciprocal space is discussed. Additionally, the difference between the eigenvalue method and the multislice method is introduced. The difference between the use of the Fourier transformation vs. the convolution in reciprocal space on multislice calculations is also discussed. Multislice computer simulations were carried out using the commercial program Cerius2 for rutile TiO2. Calculated results are compared with the results of other programs reported in the literature. The effect of slice thickness of the multislice calculation is also shown.

Fourier Analysis↗
PubMed1998

Gradient characterization using a Fourier-transform technique.

This paper describes a technique for characterizing the gradient subsystem of a magnetic resonance (MR) system. The technique uses a Fourier-transform analysis to directly measure the k-space trajectory produced by an arbitrary gradient waveform. In addition, the method can be easily extended to multiple dimensions and can be adapted to measuring residual gradient effects such as eddy currents. Several examples of gradient waveform and eddy-current measurements are presented. Also, it is demonstrated how the eddy-current measurements can be parameterized with an impulse-response formalism for later use in system tuning. When compared to a peak-fitting analysis, this technique provides a more direct extraction of the k-space measurements, which reduces the possibility of analysis error. This approach also has several advantages as compared to the conventional eddy-current measurement technique, including the ability to measure very short time constant effects.

Fourier Analysis↗