Pierre Gueth

Distributions of secondary particles in proton and carbon-ion therapy: a comparison between GATE/Geant4 and FLUKA Monte Carlo codes

Monte Carlo simulations play a crucial role for in-vivo treatment monitoring based on PET and prompt gamma imaging in proton and carbon-ion therapies. The accuracy of the nuclear fragmentation models implemented in these codes might affect the quality of the treatment verification. In this paper, we investigate the nuclear models implemented in GATE/Geant4 and FLUKA by comparing the angular and energy distributions of secondary particles exiting a homogeneous target of PMMA. Comparison results were restricted to fragmentation of (16)O and (12)C. Despite the very simple target and set-up, substantial discrepancies were observed between the two codes. For instance, the number of high energy (>1 MeV) prompt gammas exiting the target was about twice as large with GATE/Geant4 than with FLUKA both for proton and carbon ion beams. Such differences were not observed for the predicted annihilation photon production yields, for which ratios of 1.09 and 1.20 were obtained between GATE and FLUKA for the proton beam and the carbon ion beam, respectively. For neutrons and protons, discrepancies from 14% (exiting protons-carbon ion beam) to 57% (exiting neutrons-proton beam) have been identified in production yields as well as in the energy spectra for neutrons.

Phase-based block matching applied to motion estimation with unconventional beamforming strategies

A phase-based block matching method adapted to motion estimation with unconventional beamforming strategies is presented. The unconventional beamforming technique used allows us to obtain 2-D RF images with axial and lateral modulations. Based on these images, we propose a method that uses phase images instead of amplitude images. This way of proceeding allows us to provide an analytical solution to the local displacement estimation so that no minimization of a classical cost function is used for the local estimation. For this reason, the local estimator is directly applied to signals, without the need to process a complex cross-correlation function, as is done with most of the phase shift estimators. In this paper, the method is applied to elastography. Results with simulated data show that a downsampling of axial and lateral modulated signals leads to very little change in the accuracy and in the spatial resolution of the proposed method. For example, for decimation factors of 2 in the axial direction and of 4 in the lateral direction, the mean axial absolute error is 3 mum. The same estimation with original images provides a mean axial error of 0.7 mum. The accuracy of the lateral motion is unchanged in this case. The accuracy of our method with downsampled signals is an important issue in the purpose of a real-time implementation. With experimental data, for the same level of estimation error, classical block matching using the maximum of cross correlation as a local estimator requires images that are 36 times larger (in number of pixels) and consequently a computational time roughly 10 times longer. Our phase block matching is also shown to provide 10 percent less error than a motion estimation method based on seeking the zero of the complex correlation function phase. Finally, it is shown that given the separability of the local estimator that we propose, our method can be applied on both n-D signals and classical RF ultrasound images. The phase block matching method presented was implemented in real time on an ultrasound research scanner.

Transverse oscillations for tissue motion estimation.

This paper gives an overview of the methods developed for tissue motion estimation using transverse oscillation images (TO). TO images are specific radiofrequency ultrasound images featuring oscillations in both spatial directions. The initial studies on TO were published in the late 1990s. This paper reviews the main ideas and applications behind this motion estimation approach. First the origin and motivation of TO is briefly reviewed. Then the beamforming methods that lead to TO images are given, detailing the receive-only approach and the transmit-and-receive approach using synthetic aperture data. The different medical applications where TO has been used are discussed (blood flow, elastography and echocardiography), showing how it can improve motion estimation. Finally, the future perspectives of TO are outlined.

Monte Carlo simulation on heterogeneous distributed systems: A computing framework with parallel merging and checkpointing strategies

This paper introduces an end-to-end framework for efficient computing and merging of Monte Carlo simulations on heterogeneous distributed systems. Simulations are parallelized using a dynamic load-balancing approach and multiple parallel mergers. Checkpointing is used to improve reliability and to enable incremental results merging from partial results. A model is proposed to analyze the behavior of the proposed framework and help tune its parameters. Experimental results obtained on a production grid infrastructure show that the model fits the real makespan with a relative error of maximum 10%, that using multiple parallel mergers reduces the makespan by 40% on average, that checkpointing enables the completion of very long simulations and that it can be used without penalizing the makespan.

Lateral RF image synthesis using a synthetic aperture imaging technique

The oscillating profile naturally present in ultrasound images has been shown to be extremely valuable in different applications, particularly in motion estimation. Recent studies have shown that it is possible to produce images with transverse oscillations (TOs) based on a specific type of beamforming. However, there is still a great difference between the nature of the lateral oscillations produced with current methods and the axial profile of ultrasound images. In this study, we propose to combine synthetic aperture imaging (synthetic transmit aperture, STA) using a specific beamformer in both transmit mode and receive mode combined with a heterodyning demodulation method to produce lateral radiofrequency signals (LRFs). The aim was to produce lateral signals as close as possible to conventional axial signals, which would make it possible to estimate lateral displacements with the same accuracy as in the axial direction. The feasibility of this approach was validated in simulation and experimentally on an ultrasound research platform, the Ultrasonix RP system. We show that the combination of STA and the heterodyning demodulation can divide the wavelength of the LRF signals by 4 and divide the width of the lateral envelope of the point spread function (PSF) by 2 compared with the previous approaches using beamforming in receive mode only. Finally, we also illustrate the potential of our beamforming for motion estimation compared with previous TO methods.

Two-dimensional least-squares estimation for motion tracking in ultrasound elastography

This paper proposes a method of 2-D translations estimation using an a priori signal model. Two analytical signals defined with multidimensional Hilbert transform are considered and shown to have linear phases with respect to the translations to estimate. A least squares estimator (LSE) is then developed to adjust the measured phases of the complex signals to their theoretical forms. Moreover, the LSE provides an analytical solution to the 2-D translation estimation problem. The estimator is then included in a block matching method for motion tracking with ultrasound images. We compared our results with those obtained with a classical sum of absolute differences (SAD) cost function. We show that with our method there is no need of interpolating the images. Thus, for images at the original resolution level, the results obtained with the proposed estimator are largely more accurate than with SAD. Moreover, we show that using SAD on images with resolution five times higher provide roughly the same results as with our method, but the processing time is ten times higher in this case.

P4B-2 Beamforming Techniques for Motion Estimation in Ultrasound Elastography

In this paper, we will discuss about the interest and performance of synthetic aperture with transverse oscillations beamforming technique from the displacement estimation point of view. After the experimental validation of the combination of transmit and receive beamforming for transverse oscillations methods using synthetic aperture imaging performances in terms of displacement field accuracy are evaluated experimentally. Motion is estimated using a specific motion estimation method based on phase plane fittingThe quality of the estimation is evaluated by block wise correlation between the pre-deformation image and the registered post-deformation image. Results show that transmit and receive beamforming allow to increase by nearly a factor of 2 the transverse oscillations frequency increasing as well the transverse resolution by a factor of radic2 compared to receive beamforming only. This leads naturally to an increase in estimated displacement field accuracy.

Ultrasound Fourier slice imaging: a novel approach for ultrafast imaging technique

Ultrafast imaging based on plane-wave (PW) has become an intense area of research thanks to its capability of reaching frame rate higher than a thousand of frames per second. Several proposed approaches are based on Fourier-domain reconstruction. In these techniques, the Fourier transform of the received echoes is projected to the k-space corresponding to the Fourier transform of the object function. For one emitted PW, N lines along the kz axis direction are reconstructed in the k-space. We propose in this study a new acquisition scheme which allows acquiring the non-null part of the ultrasound spectrum with finer resolution. We show that this strategy allows obtaining images with slightly better lateral resolution and higher contrast-to-noise ratio (CNR) when compared to other Fourier-based techniques.

Improved resolution for ultrasound Fourier imaging

Ultrasound Fourier imaging was first initiated by Jian-yu Lu during the 90s. Using this method, one can compute an ultrasound image using a single emission, allowing a very high frame rate (up to 10000 frames per second). The main limitation of this method is the presence of geometrical artifacts, which tend to reduce the image resolution. In this study, we propose to use steered plane waves in reception instead of spherical waves as in [1]. This helps sampling the non-null part of the ultrasound spectrum with finer resolution, while reducing the presence of artifacts. A plane wave is emitted. Back-scattered signals are measured using multiple steered plane waves. Those signals 1D spectrum contain spatial information merged to create the 2D ultrasound image spectrum. Due to axial modulation, this 2D spectrum features 2 symmetric lobes. The image resolution increases when using plane waves instead of spherical waves because the sampling of these lobes is easier. The PSF size is measured both in simulation and experimentally for each method. For conventional imaging technique, this size is 155 by 311µm in axial and transverse directions. Standard Fourier imaging leads to a size of 156 by 279µm. With our spectral imaging, this size is 154µm in axial direction and 252µm in transverse direction. Using steered plane wave in a Fourier imaging framework increases the beamformed image resolution, especially in the transverse direction when compared to other beamforming methods.

Piecewise smooth reconstruction of normal vector field on digital data

We propose a novel method to regularize a normal vector field defined on a digital surface (boundary of a set of voxels). When the digital surface is a digitization of a piecewise smooth manifold, our method localizes sharp features (edges) while regularizing the input normal vector field at the same time. It relies on the optimisation of a variant of the Ambrosio‐Tortorelli functional, originally defined for denoising and contour extraction in image processing [ AT90 ]. We reformulate this functional to digital surface processing thanks to discrete calculus operators. Experiments show that the output normal field is very robust to digitization artifacts or noise, and also fairly independent of the sampling resolution. The method allows the user to choose independently the amount of smoothing and the length of the set of discontinuities. Sharp and vanishing features are correctly delineated even on extremely damaged data. Finally, our method can be used to enhance considerably the output of state‐of‐the‐art normal field estimators like Voronoi Covariance Measure [ MOG11 ] or Randomized Hough Transform [ BM12 ].