Ultrasonic strain imaging and reconstructive elastography for biological tissue

Mechanical properties of biological tissue represent important diagnostic information and are of histological and pathological relevance. In order to obtain non-invasively mechanical properties of tissue, we developed a real-time strain imaging system for clinical applications. The output data of this system also allow an inverse elastography approach leading to the spatial distribution of the relative elastic modulus of tissue. The internal displacement field of biological tissue is determined using the above mentioned strain imaging system by applying quasi-static compression to the considered tissue. Axial displacements are calculated by comparing echo signal sets obtained prior to and immediately following less than 0.1% compression, using the fast root seeking technique. Strain images representing mechanical tissue properties in a non-quantitative manner are displayed in real-time mode. For additional quantitative imaging, the stiffness distribution is calculated from the displacement field assuming the investigated material to be elastic, isotropic, and nearly incompressible. Different inverse problem approaches for calculating the shear modulus distribution using the internal displacement field have been implemented and compared. The results of an ongoing clinical study with more than 200 patients show, that our real-time strain imaging system is able to differentiate malignant and benign tissue areas in the prostate with a high degree of accuracy (sensitivity=76% and specificity=89%). The reconstruction approaches applied to the strain image data deliver quantitative tissue information and seem promising for an additional differential diagnosis of lesions in biological tissue. Our real-time system has the potential of improving diagnosis of prostate and breast cancer.     

Angle matching in intravascular elastography

Intravascular elastography is a new technique to obtain mechanical properties of the vessel wall and plaque. Mechanical information of vascular tissue is important for characterisation of different plaque components, detection of plaque vulnerability and thus choosing the proper interventional technique. The feasibility of the technique is investigated using phantoms and diseased human arteries. These studies demonstrated that elastography reveals information that is unavailable or inconclusive from the echogram alone. The technique is based on the principle that tissue strain is directly related to its mechanical properties. In intravascular elastography, the tissue is compressed using different intravascular pressures. The strain is determined using cross-correlation techniques of the radio frequency (r.f.) signals. Reliable strain estimates are only obtained when signals of corresponding tissue are correlated. Owing to catheter motion, off-centre position and non-uniform rotation of the intravascular transducer, the r.f. traces at low and at high pressure may be misaligned. Four algorithms are tested to track the corresponding ultrasound signals. Three methods (l1norm, l2norm and cross-correlation) are applied on the r.f. signal and one (l1norm) on the envelope (speckle tracking). Simulations are performed to obtain a data set with a priori knowledge of the scattering particles positions in the tissue at high and low pressure. Different positions of the catheter in the lumen, compression levels of the material and signal-to-noise ratios (SNRs) are investigated. Finally, these findings are corroborated with a phantom experiment in a water tank. From the simulations, it can be concluded that the speckle tracking algorithm has the best performance, under all circumstances. The performance decreases with larger eccentricity of the catheter and larger compression of the material. The SNR is only of minor influence. The speckle tracking algorithm has also the best performance in the phantom experiment. The performance of the speckle tracking algorithm is better than the three r.f.-based algorithms. For intravascular elastography, implementation of this method may improve the quality of the elastogram.     

Vascular tissue characterisation with IVUS elastography

Knowledge about the mechanical properties of the vessel wall and plaque is important for guiding intravascular interventional procedures and detection of plaque vulnerability. Rupture of atherosclerotic plaques is associated with acute myocardial infarction and unstable angina pectoris. In a plaque with a lipid core, the stress due to the arterial pulsation will be concentrated in the cap and a thin cap may be unable to bear this stress. In this study, the potential of intravascular elastography to characterise fibrous, fibro-fatty and fatty tissue based on their mechanical properties was investigated. Using a custom-made set-up, intravascular echograms and elastograms of excised human femoral arteries were determined. High frequency r.f. data (30 MHz) were acquired using an intravascular catheter. The tissue was compressed using intravascular pressures of 80 and 100 mmHg. The cross-sections of interest were marked with a needle for matching with histology. Using cross-correlation estimation of gated echosignals, elastograms (images of the local strain) were determined. After the intravascular experiments, the specimens were fixed in formaldehyde and processed for paraffin embedding. Sections were stained with picrosirius red and alpha-actin to counterstain collagen and smooth muscle cells (SMC), respectively. Results of vessel cross-sections with fibrous and fatty plaque regions will be presented. The elastograms of these specimens show that the strain in fatty tissue is higher than the strain in fibrous material. In conclusion, these in vitro experiments on human femoral arteries indicate the potential of intravascular elastography to characterise different plaque components.     

Measuring of viscoelastic properties of homogeneous soft solid using transient elastography: an inverse problem approach

Two main questions are at the center of this paper. The first one concerns the choice of a rheological model in the frequency range of transient elastography, sonoelasticity or NMR elastography for soft solids (20-1000 Hz). Transient elastography experiments based on plane shear waves that propagate in an Agar-gelatin phantom or in bovine muscles enable one to quantify their viscoelastic properties. The comparison of these experimental results to the prediction of the two simplest rheological models indicate clearly that Voigt's model is the better. The second question studied in the paper deals with the feasibility of quantitative viscosity mapping using inverse problem algorithm. In the ideal situation where plane shear waves propagate in a sample, a simple inverse problem based on the Helmholtz equation correctly retrieves both elasticity and viscosity. In a more realistic situation with nonplane shear waves, this simple approach fails. Nevertheless, it is shown that quantitative viscosity mapping is still possible if one uses an appropriate inverse problem that fully takes into account diffraction in solids.     

Ultrasonic elastography using sector scan imaging and a radial compression

Elastography is an imaging technique based on strain estimation in soft tissues under quasi-static compression. The stress is usually created by a compression plate, and the target is imaged by an ultrasonic linear array. This configuration is used for breast elastography, and has been investigated both theoretically and experimentally. Phenomena such as strain decay with tissue depth and strain concentrations have been reported. However in some in vivo situations, like prostate or blood vessels imaging, this set-up cannot be used. We propose a device to acquire in vivo elastograms of the prostate. The compression is applied by inflating a balloon that covers a transrectal sector probe. The 1D algorithm used to calculate the radial strain fails if the center of the imaging probe does not correspond to the center of the compressor. Therefore, experimental elastograms are calculated with a 2D algorithm that accounts for tangential displacements of the tissue. In this article, in order to gain a better understanding of the image formation process, the use of ultrasonic sector scans to image the radial compression of a target is investigated. Elastograms of homogeneous phantoms are presented, and compared with simulated images. Both show a strain decay with tissue depth. Then experimental and simulated elastograms of a phantom that contains a hard inclusion are presented, showing that strain concentrations occur as well. A method to compensate for strain decay and therefore to increase the contrast of the strain elastograms is proposed. It is expected that such information will help to interpret and possibly improve the elastograms obtained via radial compression.     

Intravascular ultrasound elastography: an overview

The composition and morphology of the atherosclerotic lesion are currently considered more important determinants of acute coronary ischemic syndromes than the degree of stenosis. When a lesion is unstable, it may rupture and cause an acute thrombotic reaction. A rupture prone plaque contains a large lipid pool covered by a thin fibrous cap. The stress in the cap increased with decreasing thickness. Additionally, it may be weakened by macrophage infiltration. Intravascular ultrasound elastography might be an ideal technique to assess the presence of lipid pools and identify high stress regions. Elastography is a technique to assess local mechanical properties of tissue. The underlying principle is that the deformation of tissue by a mechanical excitation is a function of its mechanical properties. The deformation of the tissue is determined using ultrasound. For intravascular purposes, the intraluminal pressure is used as the excitation force. The radial strain in the tissue is obtained by cross-correlation techniques on the radio frequency (rf) signal. The strain is colour-coded and plotted as a complimentary image to the IVUS echogram. Elastography was validated in vitro using diseased human coronary and femoral arteries. After the ultrasound experiments, the specimens were processed for routine histology to counterstain collagen, smooth-muscle cells, and macrophage activity. Regions were segmented in the elastograms based on their strain values. Next, the dominant plaque type (fibrous, fibro-fatty or fatty) was defined by observers blinded to the elastographic result. These experiments demonstrate that the strain in the three plaque types is different (Kruskall-Wallis p < 0.001). Especially between fibrous and fatty tissue, a highly significant difference (Wilcoxon p < 0.001) was found. In vivo, the technique is validated in an atherosclerotic Yucatan mini-pig animal model. High-resolution echo frames (30 frames per second) were acquired near end-diastole. In this phase of the pressure cycle, catheter motion was minimal. Frames with a pressure difference of approx. 5 mm Hg were taken to determine the elastograms. This in vivo validation study in Yucatan mini-pigs revealed higher strain values in fatty material (ANOVA p < 0.001). All cross-sections with a fatty plaque were identified with the elastogram by the presence of high strain values. Additionally, data are acquired in patients referred for Percutaneous Transluminal Coronary Angioplasty with the same set-up as tested in the animal study. Ultrasound data of soft, fibrous, calcified and stented plaques are acquired near end-diastole. The elastogram of soft plaques. as identified from the deformation during the pressure cycle, reveals strain values of 1% with increased strain up to 2% at the shoulders of the plaque. Calcified material, as identified from the echogram, shows low strain values of 0-0.2%. The elastogram of stented plaques reveals very low strain values, except for two regions: these are between the stent struts and at the shoulders of the plaque. In conclusion, intravascular elastography appears to be a unique tool to determine local mechanical properties in atherosclerotic lesions to identify fibrous and fatty tissue. Experiments have demonstrated the feasibility of this technique to be applied in vivo.     

Coherent cross-polarization theory for a spin-12 coupled to a general object

This work describes a segmented radial turbo-spin-echo technique (DW-rTSE) for high-resolution multislice diffusion-weighted imaging and quantitative ADC mapping. Diffusion-weighted images with an in-plane resolution of 700 microm and almost free of bulk motion can be obtained in vivo without cardiac gating. However, eddy currents and pulsatile brain motion cause severe artifacts when strong diffusion weighting is applied. This work explains in detail the artifacts in projection reconstruction (PR) imaging arising from eddy currents and describes an effective eddy current compensation based on the adjustment of gradient timing. Application of the diffusion gradients in all three orthogonal directions is possible without degradation of the images due to eddy current artifacts, allowing studies of the diffusional anisotropy. Finally, a self-navigation approach is proposed to reduce residual nonrigid body motion artifacts. Five healthy volunteers were examined to show the feasibility of this method.     

Deformation behavior of silver submicrometer-pillars prepared by nanoimprinting

Scaling in material properties is of great importance in microsystems and microelectronics where the device dimensions continuously shrink. Recently, compression tests of micron-sized pillars produced using a focused-ion beam (FIB) have become standard in the investigation of scaling effects. The influence of the fabrication process on the mechanical properties of the samples has, however, not been conclusively resolved. In this study, 130 silver pillars were fabricated using a novel embossing technique that does not pose the issues associated with FIB milling, i.e. surface amorphization and gallium contamination. Displacement-controlled compression tests on pillars with diameters of 130–3000 nm reveal for submicrometer samples an inverse proportionality of flow strength to diameter, which is associated with a wide strength distribution and a deformation taking place in large discrete strain bursts. The largest pillars show instead near bulk-like behavior. Unlike studies on other fcc materials, mechanical twinning is also an important deformation mechanism in Ag due to its very low stacking fault energy. In addition, occasional preexisting twin boundaries constituted preferential planes for localized deformation if favorably sloped. Despite the absence of ion-induced artifacts, scaling laws and absolute values of strength are comparable to those obtained via FIB milling.     

A statistical iteration approach with energy minimization to sinogram noise reduction for low-dose X-ray CT

Low-dose CT imaging has been particularly used in modern medical practice for its advantage on reducing the radiation dose to patients. However, excessive quantum noise is present in low dose X-ray imaging along with the decrease of the radiation dose; thus, there are obvious streak-like artifacts in reconstructed images. The statistical iterative reconstruction approach applied to the noisy sinogram before a filtered back-projection (FBP) is a resolution to deal with the noisy problem. In this paper, the statistical property of the noise sinogram was considered to achieve a satisfactory image reconstruction and a statistical iterative method with energy minimization was proposed to address the problem of streak-like artifacts. Simulations were performed and indicated that the proposed method could suppress noise and obviously decrease streak-like artifacts in reconstructed images.     

Correlation analysis of the beam angle dependence for elastography

Signal decorrelation is a major source of error in the displacements estimated using correlation techniques for elastographic imaging. Previous papers have addressed the variation in the correlation coefficient as a function of the applied compression for a finite window size and an insonification angle of zero degrees. The recent use of angular beam-steered radio-frequency echo signals for spatial angular compounding and shear strain estimation have demonstrated the need for understanding signal decorrelation artifacts for data acquired at different beam angles. In this paper, we provide both numerical and closed form theoretical solutions of the correlation between pre- and post-compression radio-frequency echo signals acquired at a specified beam angle. The expression for the correlation coefficient obtained is a function of the beam angle and the applied compression for a finite duration window. Accuracy of the theoretical results is verified using tissue-mimicking phantom experiments on a uniformly elastic phantom using beam-steered data acquisitions on a linear array transducer. The theory predicts a faster decorrelation with changes in the beam or insonification angle for longer radio-frequency echo signal segments and at deeper locations in the medium. Theoretical results provide useful information for improving angular compounding and shear strain estimation techniques for elastography.     

Accelerating MRI Using GROG Gridding Followed by ESPIRiT for Non-Cartesian Trajectories

Parallel imaging plays an important role to reduce data acquisition time in magnetic resonance imaging (MRI). Under-sampled non-Cartesian trajectories accelerate the MRI scan time, but the resulting images may have aliasing artifacts. To remove these artifacts, a variety of methods have been developed within the scope of parallel imaging in the recent past. In this paper, the use of Eigen-vector-based iterative Self-consistent Parallel Imaging Reconstruction Technique (ESPIRiT) along with self-calibrated GRAPPA operator gridding (self-calibrated GROG) on radial k-space data for accelerated MR image reconstruction is presented. The proposed method reconstructs the solution image from non-Cartesian k-space data in two steps: First, the acquired radial data is gridded using self-calibrated GROG and then ESPIRIT is applied on this gridded data to get the un-aliased image. The proposed method is tested on human head data and the short-axis cardiac radial data. The quality of the reconstructed images is evaluated using artifact power (AP), root-mean-square error (RMSE) and peak signal-to-noise ratio (PSNR) at different acceleration factors (AF). The results of the proposed method (GROG followed by ESPIRiT) are compared with GROG followed by pseudo-Cartesian GRAPPA reconstruction approach (conventionally used). The results show that the proposed method provides considerable improvement in the reconstructed images as compared to conventionally used pseudo-Cartesian GRAPPA with GROG, e.g., 87, 67 and 82% improvement in terms of AP for 1.5T, 3T human head and short-axis cardiac radial data, 63, 45 and 57% improvement in terms of RMSE for 1.5T, 3T human head and short-axis cardiac radial data, 11, 7 and 9% improvement in terms of PSNR for 1.5T, 3T human head and short-axis cardiac radial data, respectively, at AF = 4.     

The adaptive sinogram restoration algorithm based on anisotropic diffusion by energy minimization for low-dose X-ray CT

Apparent streak artifacts will present in reconstructed images due to excessive quantum noise in low-dose X-ray imaging process. Estimating a noise-free sinogram to satisfy the filtered back-projection (FBP) reconstruction is an effective way to solve this problem. In this paper, we propose a novel sinogram noise reduction method by energy minimization. An adaptive smoothness parameter based on a modified anisotropic diffusion coefficient is applied for an optimal estimation. The smoothness parameter can make the method effectively adjust the degree of smoothness according to the noise level and the region feature in the sinogram. Visual effect together with quantitative analysis of the experimental result shows the developed approach has the excellent performance in protection of the edge and removal of streak artifacts in the reconstructed image.     

Quo vadis elasticity imaging?

In the past decade, an important field that has emerged as complementary to ultrasonic imaging is that of elasticity imaging. The term encompasses a variety of techniques that can depict a mechanical response or property of tissues. In ultrasound, its premise is built on two important facts: (a) that significant differences between mechanical properties of several tissue components exist and (b) that the information contained in the coherent scattering, or speckle, is sufficient to depict these differences following an external or internal mechanical stimulus. Parameters, such as velocity of vibration, displacement, strain, strain rate, velocity of wave propagation and elastic modulus, have all been demonstrated feasible in their estimation and have resulted in the accurate depiction of stiffer tissue masses, such as tumors, high-intensity focused ultrasound (HIFU) lesions and atherosclerotic plaques. More recently, through the development of ultrafast algorithms tailored to suitable hardware as well as the familiarity of the physician with the sensitivity of the methods used, one elasticity imaging technique in particular, elastography, has been shown applicable in a typical clinical ultrasound setting. In other words, elastograms can currently be obtained at quasi real-time (approximately at a frame rate of 8 frames/s) and with the use of a hand-held transducer (as opposed to the previously used frame-suspended setup) during and simultaneously with an ultrasound exam of, e.g., the breast or the prostate. The higher frame rate available with certain clinical ultrasound scanners has also resulted in the successful application of elasticity imaging techniques on the myocardium and monitoring its deformation over several cardiac cycles for the detection of ischemic regions. As a result, elasticity imaging with its ever increasing number of applications and demonstrated applicability in a typical, clinical ultrasound setting promises to make an important contribution to the ultrasound practice as we know it.     

Improved depiction of small anatomic structures in MR images using Gaussian-weighted spirals and zero-filled interpolation

Partial-volume artifacts reduce the contrast and continuity of small structures in magnetic resonance images. Zero-filled interpolation (ZFI) has been known for some time as a useful technique to reduce partial-volume artifacts and improve the appearance of small structures and edges. However, its use is limited by the fact that ZFI can exacerbate image artifacts. For example, it can exacerbate Gibbs ringing, also known as the truncation artifact, which manifests itself as spurious ringing around sharp edges. Currently, the most common technique to address this problem is post-acquisition filtering, which causes blurring in the image. Using ZFI in conjunction with a variable-density sampling method designed to reduce ringing is proposed as a possible solution to this problem. This approach is demonstrated with a Gaussian-weighted spiral and is compared to conventional spiral sampling both with and without the application of a filter used to reduce ringing. The two spiral sampling techniques are compared using simulations, phantom images, and in vivo brain images. The Gaussian-weighted spiral demonstrates reduced ringing without the loss of spatial resolution commonly associated with post-acquisition filtering. Additionally, this sampling technique is shown to work well in conjunction with ZFI to reduce partial-volume artifacts without the apparent increase in Gibbs ringing usually associated with zero-filled reconstruction. This approach will be most useful for imaging techniques such as MR angiography which are known to be sensitive to partial-volume effects, as well as when imaging anatomic regions associated with more severe Gibbs ringing.     

O-space with high resolution readouts outperforms radial imaging

PurposeWhile O-Space imaging is well known to accelerate image acquisition beyond traditional Cartesian sampling, its advantages compared to undersampled radial imaging, the linear trajectory most akin to O-Space imaging, have not been detailed. In addition, previous studies have focused on ultrafast imaging with very high acceleration factors and relatively low resolution. The purpose of this work is to directly compare O-Space and radial imaging in their potential to deliver highly undersampled images of high resolution and minimal artifacts, as needed for diagnostic applications. We report that the greatest advantages to O-Space imaging are observed with extended data acquisition readouts.Theory and methodsA sampling strategy that uses high resolution readouts is presented and applied to compare the potential of radial and O-Space sequences to generate high resolution images at high undersampling factors. Simulations and phantom studies were performed to investigate whether use of extended readout windows in O-Space imaging would increase k-space sampling and improve image quality, compared to radial imaging.ResultsExperimental O-Space images acquired with high resolution readouts show fewer artifacts and greater sharpness than radial imaging with equivalent scan parameters. Radial images taken with longer readouts show stronger undersampling artifacts, which can cause small or subtle image features to disappear. These features are preserved in a comparable O-Space image.ConclusionsHigh resolution O-Space imaging yields highly undersampled images of high resolution and minimal artifacts. The additional nonlinear gradient field improves image quality beyond conventional radial imaging.     

Deformation textures produced in diamond anvil experiments, analysed in radial diffraction geometry

Diamond anvil cells may not only impose pressure upon a sample but also a compressive stress that produces elastic and plastic deformation of polycrystalline samples. The plastic deformation may result in texture development if the material deforms by slip or mechanical twinning, or if grains have a non-equiaxed shape. In radial diffraction geometry, texture is revealed by variation of intensity along Debye rings relative to the compression direction. Diffraction images (obtained by CCD or image plate) can be used to extract quantitative texture information. Currently the most elegant and powerful method is a modified Rietveld technique as implemented in the software package MAUD. From texture data one can evaluate the homogeneity of strain in a diamond anvil cell, the strain magnitude and deformation mechanisms, the latter by comparing observed texture patterns with results from polycrystal plasticity simulations. Some examples such as olivine, magnesiowuestite, MgSiO(3) perovskite and ε-iron are discussed.     

Finite element modeling and intravascular ultrasound elastography of vulnerable plaques: parameter variation

BACKGROUND AND GOAL: More than 60% of all myocardial infarction is caused by rupture of a vulnerable plaque. A vulnerable plaque can be described as a large, soft lipid pool covered by a thin fibrous cap. Plaque material composition, geometry, and inflammation caused by infiltration of macrophages are considered as major determinants for plaque rupture. For diagnostic purposes, these determinants may be obtained from elastograms (i.e. radial strain images), which are derived from intravascular ultrasound (IVUS) measurements. IVUS elastograms, however, cannot be interpreted directly as tissue component images, because radial strain depends upon plaque geometry, plaque material properties, and used catheter position. To understand and quantify the influence of these parameters upon measured IVUS elastograms, they were varied in a finite element model (FEM) that simulates IVUS elastograms of vulnerable plaques. MATERIALS AND METHODS: IVUS elastography measurements were performed on a vessel mimicking phantom, with a soft plaque embedded in a hard wall, and an atherosclerotic human coronary artery containing a vulnerable plaque. Next, FEMs were created to simulate IVUS elastograms of the same objects. In these FEMs the following parameters were varied: Young's modulus (E), Poisson's ratio (nu) in range 0.49-0.4999, catheter position (translation of 0.8 mm), and cap thickness (t) in range 50-350 microm. Hereby the resulting peak radial strain (PRS) was determined and visualized. RESULTS: Measured static E for phantom was 4.2 kPa for plaque and 16.8 kPa for wall.Variation of E-wall in range 8.4-33.2 kPa and/or E-plaque in range 2.1-8.4 kPa using the phantom FEM, gave a PRS variation of 1.6%, i.e. from 1.7% up to almost 3.3%; for variation in nu this was only 0.07%, i.e. from 2.37% up to 2.44%. Variation of E-lipid in range 6.25-400 kPa and E-cap in range 700-2300 kPa using the artery FEM, gave a PRS variation of 3.1%, i.e. from 0.6% up to 3.7%. The PRS was higher for lower E-lipid and E-cap; it was located at a shoulder of the lipid pool. Variation of nu gave only a variation of 0.17%. Variation of t and E-cap resulted in a PRS variation of 1.4%, i.e. from 0.3% up to 1.7%; thinner and weaker caps gave higher PRS. Catheter position variation changed radial strain value. CONCLUSIONS: Measured IVUS elastograms of vulnerable plaques depend highly upon the Young's modulus of lipid and cap, but not upon the Poisson's ratio. Different catheter positions result in different IVUS elastograms, but the diagnostically important high strain regions at the lipid shoulders are often still detectable. PRS increases when cap weakens or cap thickness decreases.     

Optical strain gauge vs. traditional strain gauges for concrete elasticity modulus determination

In the field of Civil Engineering, the evaluation of concrete elasticity modulus relies heavily on traditional strain gauges like electric resistance strain gauge, demec mechanical strain and compressometer. These strain gauges work pretty well but the sample surface preparation and the strain gauge adhesion are tedious and time-consuming. On the other hand, convenient non-destructive testing methods like rebound hammer and ultrasonic pulse velocity have the disadvantage of large data scattering and low reliability. In this paper, we aim to develop a handy, reliable and cost-effective method for concrete property evaluation based on an optical technique called automatic digital image correlation (ADIC), which acts as an optical strain gauge. By incorporating an automatic searching approach, the proposed ADIC greatly expands the convergence range of normal digital image correlation (DIC) algorithms and is able to automatically determine concrete strain in real time without human interaction. After the optical strain gauge is set up, the axial concrete strain as well as the stress-strain curve is plotted in real time when the concrete sample under testing is loaded. The concrete elasticity modulus is then determined from the curve. A series of experiments have been conducted to evaluate the performance of the proposed optical strain gauge. Comparisons are also carried out between the proposed optical strain gauge and traditional strain gauges such as electronic strain gauge, compressometer and demec mechanical strain gauge, which further confirm the accuracy, convenience and reliability of the proposed technique.     

Automated artifact detection and removal for improved tensor estimation in motion-corrupted DTI data sets using the combination of local binary patterns and 2D partial least squares

Signal variation in diffusion-weighted images (DWIs) is influenced both by thermal noise and by spatially and temporally varying artifacts, such as rigid-body motion and cardiac pulsation. Motion artifacts are particularly prevalent when scanning difficult patient populations, such as human infants. Although some motion during data acquisition can be corrected using image coregistration procedures, frequently individual DWIs are corrupted beyond repair by sudden, large amplitude motion either within or outside of the imaging plane. We propose a novel approach to identify and reject outlier images automatically using local binary patterns (LBP) and 2D partial least square (2D-PLS) to estimate diffusion tensors robustly. This method uses an enhanced LBP algorithm to extract texture features from a local texture feature of the image matrix from the DWI data. Because the images have been transformed to local texture matrices, we are able to extract discriminating information that identifies outliers in the data set by extending a traditional one-dimensional PLS algorithm to a two-dimension operator. The class-membership matrix in this 2D-PLS algorithm is adapted to process samples that are image matrix, and the membership matrix thus represents varying degrees of importance of local information within the images. We also derive the analytic form of the generalized inverse of the class-membership matrix. We show that this method can effectively extract local features from brain images obtained from a large sample of human infants to identify images that are outliers in their textural features, permitting their exclusion from further processing when estimating tensors using the DWIs. This technique is shown to be superior in performance when compared with visual inspection and other common methods to address motion-related artifacts in DWI data. This technique is applicable to correct motion artifact in other magnetic resonance imaging (MRI) techniques (e.g., the bootstrapping estimation) that use univariate or multivariate regression methods to fit MRI data to a pre-specified model.     

Experimental three dimensional strain estimation from ultrasonic sectorial data

Most of the studies devoted to elastography are focused on the estimation of the axial component of the strain. However when subjected to any load, whatever the direction, soft biological media deform in the three spatial dimensions. The aim of our work is to build a three dimensional strain mapping from data acquired with a 3D clinical sectorial probe. The estimation of radial strain is based on the estimation of local scaling factors. A method of cross-correlation of interpolated signals between adjacent radiofrequency lines was used to estimate the angular displacement and strain. For the sectorial strain estimation, the same displacement estimation technique has been implemented. The method has been tested on experimental data acquired on calibrated phantoms and compared to simulation.