No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Abstract
Freehand three-dimensional (3D) ultrasound imaging is an important medical imaging modality in computer-assisted clinical diagnosis and image-guided intervention. In this paper, we present a novel Bayesian-based nonlocal method for the accurate volume reconstruction of freehand 3D ultrasound imaging with irregularly spaced B-scans. In the algorithm, each pixel is represented as the Gamma distribution which corresponds to the speckle noise generated by the interaction of the acoustic wave with the tissues. The variational reconstruction functional is associated with a nonlocal denoising term and a nonlocal inpainting term. To suppress speckle noise in the ultrasound image, the observed data is filtered via nonlocal total variation method firstly. The nonlocal denoising model is adapted to the speckle noise by substituting the Pearson distance-based weight function for the Gaussian weight function. To interpolate the missing data, a new inpainting scheme derived from the nonlocal means filter and its implementation based on fast marching method are introduced to fill the empty regions. This makes interpolation of missing data more accurate and effective. The Pearson distance function derived from the Bayesian estimator is not only used for speckle reduction, but also serves as weight function for building nonlocal means-based inpainting algorithm. Experimental results on synthetic cube data, in-vitro ultrasound abdominal phantom and in-vivo liver of human subject and comparisons with some classical and recent algorithms are used to demonstrate its improvement in both speckle suppression and edge preservation in 3D ultrasound reconstruction.
... The volume or surface 3D reconstruction can provide useful clues for medical image analysis. Previous 3D imaging methods for medical images can be classified into computed tomography (CT) based [5,17,20,28,35], magnetic resonance imaging (MRI) based [9,12,19,21], and ultrasound (US) based approaches [2,13,14,26,32,33,36]. In addition, many researchers in this domain also report a number of other methods. ...
... 3D reconstruction from freely acquired 2D images usually needs position data of the 2D slices [3,10,32,36]. These reconstruction algorithms can be categorized into three types according to their implementation: voxel-based, pixel-based, and function-based. ...
... The voxel-based methods traverse all voxels in a predefined volume and insert the corresponding pixels from the 2D input images [2,11]. The voxel-based methods usually include voxel-nearest neighbor method, voxel-interpolation method, as well as voxel-distance weighted method [36]. However, when there is more speckle noise, sparse sampling data and misalignment in the original images, the reconstructed error is large in these methods. ...
Although the ultrasonic C-scan technique has been extensively applied in nondestructive testing (NDT) in recent years, 3D reconstruction from ultrasonic C-scan images has not been well addressed. This paper develops a novel and efficient 3D reconstruction technique based on an improved K-nearest neighbor filtering for ultrasonic C-scan data of the tissue-mimicking phantoms. An edge-points-predicting approach based on K-nearest neighbor filtering is first proposed to predict the undetected edge points and to reduce the noise points for 2D ultrasonic images. Then, the 3D model is reconstructed from the clean edges by utilizing the surface rendering algorithm. The proposed approach is validated using the ultrasonic C-scan data of a liver model embedded in a tissue-mimicking phantom. The comparisons with other methods are presented in the experiments. The results demonstrate the effectiveness and the significantly improved reconstruction results of the proposed approach.
... For example, many studies used transrectal or transvaginal transducers for 3-D US imaging, did not provide a detailed explanation about their method of 3-D US reconstruction, but only the 3-D volume results. Short analyses and descriptions of the remaining 45 papers (Boctor et al. 2001;Chen et al. 2009Chen et al. , 2014Coupe et al. 2007;Deng et al. 2012;Fenster et al. 2011;Gee et al. 2002Gee et al. , 2004Gilliam et al. 2006;Hassenpflug et al. 2005;Housden et al. 2006aHousden et al. , 2007Zheng 2006a, 2008;Huang et al. , 2009bHuang et al. , 2013Huang et al. , 2015Karamalis et al. 2009;Kohyama et al. 2005;Lindseth et al. 2003;MacGillivray et al. 2009;Pagoulatos et al. 2000;Penney et al. 2004;Poon and Rohling 2006;Prager et al. 2002Prager et al. , 1999Prager et al. , 2003Qiu et al. 2011;Roxborough and Nielson 2000;San Jos e-Est epar et al. 2003;Sanches and Marques 2003;Scheipers et al. 2010;Sun and Anthony 2012;Sun et al. 2013Sun et al. , 2014Toonkum et al. 2011;Treece et al. 2001a;Varandas et al. 2004;Wen et al. 2013Wen et al. , 2015Yu et al. 2006aYu et al. , 2005Zhang et al. 2002Zhang et al. , 2004 are found under Qualitative Analysis of Research Paper Studies and Quantitative Analysis. ...
... In general, the main purpose of these algorithms is to construct a 3-D US volume on a regular or irregular grid from a set of B-scans with minimum computational requirements and without damaging or losing the underlying shape of the data. Comparisons of these algorithms can be found in the literature (Miller et al. 2012;Rohling et al. 1999a;Solberg et al. 2007;Wen et al. 2015). Selection of a suitable grid for the construction of the volume is one of the main issues in volume reconstruction. ...
... Bayesian-based non-local interpolation using the gamma distribution along with the fast marching method and many denoising approaches, such as the Pearson distance function and filtering approaches, has been applied to produce high-quality volume reconstruction in a freehand method (Wen et al. 2015). A joint process of reconstruction and alignment was also used in a Bayesian framework, in a similar approach (Sanches and Marques 2003), and optimization algorithms were used to estimate the volume and alignment parameters in the 3-D volume construction. ...
Two-dimensional ultrasound (US) imaging has been successfully used in clinical applications as a low-cost, portable and non-invasive image modality for more than three decades. Recent advances in computer science and technology illustrate the promise of the 3-D US modality as a medical imaging technique that is comparable to other prevalent modalities and that overcomes certain drawbacks of 2-D US. This systematic review covers freehand 3-D US imaging between 1970 and 2017, highlighting the current trends in research fields, the research methods, the main limitations, the leading researchers, standard assessment criteria and clinical applications. Freehand 3-D US systems are more prevalent in the academic environment, whereas in clinical applications and industrial research, most studies have focused on 3-D US transducers and improvement of hardware performance. This topic is still an interesting active area for researchers, and there remain many unsolved problems to be addressed.
... This information enables the position and orientation of the sensor coils to be acquired [109]. [115] with permission. (b) An optical tracker based freehand 3D US imaging system. ...
... The preliminary results demonstrated the capability of the low-cost method for recon- Figure 7. (a) The typical configuration of a freehand 3D US imaging system. Reprinted from [115] with permission. (b) An optical tracker based freehand 3D US imaging system. ...
With the rapid advancement of tracking technologies, the applications of tracking systems in ultrasound imaging have expanded across a wide range of fields. In this review article, we discuss the basic tracking principles, system components, performance analyses, as well as the main sources of error for popular tracking technologies that are utilized in ultrasound imaging. In light of the growing demand for object tracking, this article explores both the potential and challenges associated with different tracking technologies applied to various ultrasound imaging applications, including freehand 3D ultrasound imaging, ultrasound image fusion, ultrasound-guided intervention and treatment. Recent development in tracking technology has led to increased accuracy and intuitiveness of ultrasound imaging and navigation with less reliance on operator skills, thereby benefiting the medical diagnosis and treatment. Although commercially available tracking systems are capable of achieving sub-millimeter resolution for positional tracking and sub-degree resolution for orientational tracking, such systems are subject to a number of disadvantages, including high costs and time-consuming calibration procedures. While some emerging tracking technologies are still in the research stage, their potentials have been demonstrated in terms of the compactness, light weight, and easy integration with existing standard or portable ultrasound machines.
... Hence, nonlocal regularizers can effectively model long-range dependencies and yield improvements in reconstruction results. Inspired by the success of nonlocal means (NLM) filtering for image denoising [13], many nonlocal regularization-based methods have also been proposed for various image processing applications [11,12,[14][15][16][17][18][19][20][21][22][23][24][25][26][27], and also CS image restoration [28][29][30][31][32][33][34]. ...
... However, learning the atoms from a set of training signals belonging to signal class of interest would result in dictionaries with the capability of better matching the content of the signals. It has been experimentally shown that these adaptive dictionaries outperform the non-adaptive ones in many signal processing applications such as image compression [31,51,55], denoising [19,57], deblurring [17,19], interpolation [18,21], and super-resolution [17,24,56]. ...
Compressive sensing (CS) is a recently emerging technique and an extensively studied problem in signal and image processing, which enables joint sampling and compression into a unified approach. Recently, local smoothness and nonlocal self-similarity have both led to superior sparsity priors for CS image restoration. In this paper, first, a new sparsity measure called joint adaptive sparsity measure (JASM) is introduced. The proposed JASM enforces both local sparsity and nonlocal 3D sparsity in transform domain, concurrently, providing a powerful mechanism for characterizing the structured sparsities of natural image. More precisely, the local sparsity depicts the local smoothness redundancies exploited by an adaptively learned sparsifying basis, and the nonlocal 3D sparsity corresponds to the nonlocal self-similarity constraint achieved by a new proposed nonlocal statistical sparse modeling. Then, two novel techniques for high-fidelity CS image and video recovery via JASM are proposed. The proposed methods are formulated in form of minimization functional under regularization-based framework which is solved via an efficient alternating minimization algorithm based on split Bregman framework. Comprehensive experimental results are reported to manifest the effectiveness of the proposed methods compared with the current state-of-the-art methods in CS image/video restoration.
... A Bayesian based nonlocal method was used for accurate volume reconstruction of irregular interval B-scan freehand 3D ultrasound imaging. This method uses gamma distribution instead of traditional Rayleigh distribution to achieve better suppression of speckle noise [17]. A game controller position tracker has also been used to design and manufacture a low-cost 3D ultrasound system based on standard 2D ultrasound. ...
In the medical field, 3D ultrasound reconstruction can visualize the internal structure of patients, which is very important for doctors to carry out correct analyses and diagnoses. Furthermore, medical 3D ultrasound images have been widely used in clinical disease diagnosis because they can more intuitively display the characteristics and spatial location information of the target. The traditional way to obtain 3D ultrasonic images is to use a 3D ultrasonic probe directly. Although freehand 3D ultrasound reconstruction is still in the research stage, a lot of research has recently been conducted on the freehand ultrasound reconstruction method based on wireless ultrasonic probe. In this paper, a wireless linear array probe is used to build a freehand acousto-optic positioning 3D ultrasonic imaging system. B-scan is considered the brightness scan. It is used for producing a 2D cross-section of the eye and its orbit. This system is used to collect and construct multiple 2D B-scans datasets for experiments. According to the experimental results, a freehand 3D ultrasonic reconstruction method based on depth learning is proposed, which is called sequence prediction reconstruction based on acoustic optical localization (SPRAO). SPRAO is an ultrasound reconstruction system which cannot be put into medical clinical use now. Compared with 3D reconstruction using a 3D ultrasound probe, SPRAO not only has a controllable scanning area, but also has a low cost. SPRAO solves some of the problems in the existing algorithms. Firstly, a 60 frames per second (FPS) B-scan sequence can be synthesized using a 12 FPS wireless ultrasonic probe through 2–3 acquisitions. It not only effectively reduces the requirement for the output frame rate of the ultrasonic probe, but also increases the moving speed of the wireless probe. Secondly, SPRAO analyzes the B-scans through speckle decorrelation to calibrate the acousto-optic auxiliary positioning information, while other algorithms have no solution to the cumulative error of the external auxiliary positioning device. Finally, long short-term memory (LSTM) is used to predict the spatial position and attitude of B-scans, and the calculation of pose deviation and speckle decorrelation is integrated into a 3D convolutional neural network (3DCNN). Prepare for real-time 3D reconstruction under the premise of accurate spatial pose of B-scans. At the end of this paper, SPRAO is compared with linear motion, IMU, speckle decorrelation, CNN and other methods. From the experimental results, it can be observed that the spatial pose deviation of B-scans output using SPRAO is the best of these methods.
... De manière globale, [Zhang et al., 2004] ont travaillé sur des fantômes et des tissus humains. Plus spécifiquement, des recherches ont été dirigées sur l'étude des muscles [Barber et al., 2019[Barber et al., , 2016MacGillivray et al., 2009], sur la neurologie interventionnelle [Miller et al., 2012], sur le foie [Wen et al., 2015], en obstétrique [Cai et al., 2019], sur la colonne vertébrale [Ottacher et al., 2020] ou sur les reins [Benjamin et al., 2020]. Des systèmes à moindre coût ont été développés et testés sur des fantômes de foetus en utilisant une manette PlayStation Move et la caméra PlayStation Eye correspondante [Chan et al., 2019]. ...
Les procédures de revascularisation endovasculaires périphériques sont très fréquentes chez les patients atteints d’artériopathie oblitérante des membres inférieurs (AOMI). Le bilan préopératoire comprend de manière systématique un écho-doppler, parfois complété par un angioscanner ou un angio-IRM 3D. En l’absence d’imagerie 3D, la procédure de revascularisation commence par une artériographie diagnostique complète afin d’identifier les lésions avant de les traiter. Ce travail vise à augmenter l’apport de l’échographie grâce à la réalisation d’une cartographie préopératoire complète à l’échographie. En se basant sur une sonde échographique 2D, des réseaux de deep learning ont été entrainés pour estimer le déplacement relatif entre deux images consécutives d’une séquence de l’artère fémorale pour la reconstruire en 3D. Un réseau de segmentation permet d’extraire l’artère et de dimensionner les différentes lésions (longueur, diamètres). Un premier réseau dédié à l’estimation des déplacements dans les 6 axes de direction et un deuxième focalisé sur l’axe du plan d’élévation ont été proposés et évalués. Ce dernier repose sur la segmentation artérielle pour construire un volume centré sur l’artère et générer des vues stretched offrant un fort intérêt diagnostic. L’approche proposée est une étape vers l’utilisation de l’imagerie ultrasonore pour la cartographie 3D préopératoire, afin de simplifier l’identification et le dimensionnement des lésions et ainsi réduire la toxicité des procédures de revascularisations endovasculaires périphériques.
... To reduce speckle, a Markov random field-based (MRF) filter [20] is introduced to despcekle the PNN-reconstructed volume. In [21] , a Bayesian-based nonlocal total variation method is introduced to reduce speckle noise for the acquired B-scan images before volume reconstruction. The FBMs is another important means for volume-based reconstruction. ...
Freehand three-dimensional (3D) ultrasound imaging is an attractive research area because it is capable of providing large field of view and high in-plane resolution image to allow better illustration of complex anatomy structures. However, reconstructed image is corrupted with speckle noise and artifacts in the conventional reconstructed volume data. In this paper, we propose a simple but effective adaptive kernel regression method for volume reconstruction from freehand swept B-scan images. By creating a linear model for estimating the homogeneous region of the B-scan image and learning the parameters of the model with a supervised learning method, the statistical characteristic of speckle can be well recovered. With the learned linear model of speckle, we can easily estimate the homogenous region and reconstruct image with speckle reduction and edge preservation via the adaptive turning of the smoothing parameters of the kernel regression. Our algorithm lends itself to parallel processing, and yields a 288 × speedup on a graphics processing unit (GPU). Experiments on the simulated data, ultrasonic abdominal phantom and in-vivo liver of human subject and comparisons with some classical and recent algorithms are used to demonstrate its improvements in both volume reconstruction accuracy and efficiency.
... If more than one pixel runs through the voxel, then the voxel value can be the average (Nelson and Pretorius [32], Gobbi and Peters [33]), maximum value (Nelson and Pretorius [32]), the most recent value (Ohbuchi et al. [34]), or the first value (Trobaugh et al. [30]) of the pixels. Other investigators proposed some comparatively complex but improved interpolation algorithms for more accurate imaging [35][36][37]. These methods introduce a local neighborhood called kernel around the pixel to distribute the pixel value to the contained voxels. ...
Real-time three-dimensional (3D) ultrasound (US) has attracted much more attention in medical researches because it provides interactive feedback to help clinicians acquire high-quality images as well as timely spatial information of the scanned area and hence is necessary in intraoperative ultrasound examinations. Plenty of publications have been declared to complete the real-time or near real-time visualization of 3D ultrasound using volumetric probes or the routinely used two-dimensional (2D) probes. So far, a review on how to design an interactive system with appropriate processing algorithms remains missing, resulting in the lack of systematic understanding of the relevant technology. In this article, previous and the latest work on designing a real-time or near real-time 3D ultrasound imaging system are reviewed. Specifically, the data acquisition techniques, reconstruction algorithms, volume rendering methods, and clinical applications are presented. Moreover, the advantages and disadvantages of state-of-the-art approaches are discussed in detail.
Objective:
Freehand 3D ultrasound volume reconstruction has received considerable attention in medical research because it can freely perform spatial imaging at a low cost. However, the uneven distribution of the original ultrasound images in space reduces the reconstruction effect of the traditional method.
Approach:
An adaptive tetrahedral interpolation algorithm is proposed to reconstruct 3D ultrasound volume data. The algorithm adaptively divides the unevenly distributed images into numerous tetrahedrons and interpolates the voxel value in each tetrahedron to reconstruct 3D ultrasound volume data.
Main results:
Extensive experiments on simulated and clinical data confirm that the proposed method can achieve more accurate reconstruction than six benchmark methods. Specifically, the averaged interpolation error at the gray level can be reduced by 0.22-0.82, and the peak signal-to-noise ratio and the mean structure similarity can be improved by 0.32-1.83 dB and 0.01-0.05, respectively.
Significance:
With the parallel implementation of the algorithm, one 3D ultrasound volume data with size 279×279×276 can be reconstructed from 100 slices 2D ultrasound images with size 200×200 at 1.04 seconds. Such a quick and accurate approach has practical value in medical research.
Inpainting represents a procedure which can restore the lost parts of an image based upon the residual information. We present an inpainting network that consists of an Encoder-Decoder pipeline and a multi-dimensional adversarial network. The Encoder-Decoder pipeline extracts features from the input image with missing area and learns these features. Through unsupervised learning, the pipeline can predict and fill the missing region with the most reasonable content. Meanwhile the multi-dimensional adversarial network identifies the difference between the ground truth and the generated images both in detail and in general. Compared with the traditional training procedure, our model combines with Wasserstein Distance that enhances the stability of network training. The network is training specifically on street view images and not only performs a satisfying outcome, but also shows competitiveness when comparing with existing methods.
Robotic ultrasound systems have turned into clinical use over the past decades, increasing precision and quality of medical operations. In this paper, we propose a robot-assisted and tele-controlled ultrasound scanning system for three-dimensional imaging. A six degree-of-freedom (DOF) robot arm acting as an upper-limb prosthesis remotely controlled by an operator drove the ultrasound probe to scan the skin surfaces. Via internet, the operator in a distant place monitored the robotic prosthesis and the patient through four video cameras on-site. According to the real-time video streams, the operator could manipulate a joystick for sending instructions from the remote operating terminal to the system on the patient side to adjust the probe motions. In this way, the safety and accuracy of the proposed system could be guaranteed. During the probe scanning, the acquired B-scans and their positional data were transferred to the remote workstation for the subsequent volume reconstruction and visualization. In-vitro and in-vivo experiments were conducted to validate the feasibility and performance of the proposed system. The quantitative experimental results show that the error for volume measurement was less than 1.1%, indicating that our system could accurately and flexibly control the probe scanning and produce 3D ultrasound images with high quality.
3D ultrasound provides physicians with volume rendering of patient ultrasound data. This medical ultrasound technique carries important clinical value in image-guided surgery. However, 3D ultrasound is not widely used because obtaining 3D ultrasound images is difficult and costly, and the technology used is incapable of scanning large-volume organs. This study proposes a novel method for 3D freehand ultrasound reconstruction, which can accurately restore 3D volumes from 2D B-scans through cube propagation. The 3D multi-resolution method creates a 3D volume pyramid and a 3D texture pyramid to preserve the structural and textural details of the imaged volume. Computation efficiency is improved with the 3D Approximate Nearest Neighbors (ANN) search technique, which is proposed to accelerate the nearest-neighbor search through the whole 3D ultrasound volume. Texture information in the texture pyramid is added to the cube distance in the nearest-neighbors searching step to identify suitable nearest neighbors. Experiments performed on clinical and synthetic datasets demonstrate that the proposed algorithm is robust and can achieve quality results in a short execution time relative to current reconstruction techniques.
The image inpainting quality depends on the method of patch priority calculation and best matching patch selection method. Image inpainting can be applied in repairing the damaged images and removing the unwanted objects in the images. With reference to all the existing methods of image inpainting, the patch priority calculation and best matching patch selection play a major role in getting good results. In this paper, an improved patch priority calculation method is proposed by introducing regularization factor and adaptive coefficients. The best matching patch selection is determined using two patch searching methods such as the sum of squared error (SSE) and the sum of absolute difference (SAD) distance methods. The best combination is identified with respect to and adaptive coefficients values for highest patch priority calculation. This best combination is used along with SSE and SAD patch selection method for repairing the damaged images and removing of the object in images. The results from the proposed inpainting method are compared with the existing inpainting methods. The metrics of proposed inpainting, such as peak signal to noise ratio, mean square error and time elapsed for different images are obtained and compared. The results for the several images obtained from the proposed inpainting method using MATLAB are presented in this paper to observe the quality of the images.
Robotic ultrasound systems have turned into clinical use over the past few decades, increasing precision and quality of medical operations. In this paper, we propose a fully automatic scanning system for three-dimensional (3-D) ultrasound imaging. A depth camera was first used to obtain the depth data and color data of the tissue surface. Based on the depth image, the 3-D contour of the tissue was rendered and the scan path of ultrasound probe was automatically planned. Following the scan path, a 3-D translating device drove the probe to move on the tissue surface. Simultaneously, the B-scans and their positional information were recorded for subsequent volume reconstruction. In order to stop the scanning process when the pressure on the skin exceeded a preset threshold, two force sensors were attached to the front side of the probe for force measurement. In vitro and in vivo experiments were conducted for assessing the performance of the proposed system. Quantitative results show that the error of volume measurement was less than 1%, indicating that the system is capable of automatic ultrasound scanning and 3-D imaging. It is expected that the proposed system can be well used in clinical practices.
Image inpainting restores lost or deteriorated parts of images according to the information of known regions. Criminisi has proposed an effective exemplar-based inpainting method, which has the advantages of both texture synthesis and diffusion- based inpainting. Yet, it has its own flaws of fast priority dropping and visual inconsistency. In this paper, we propose a space varying updating strategy for the confidence term and a matching confidence term to improve the filling priority estimation. We propose structure consistent patch matching to take the distribution of source and target patch differences into account. Fast Fourier transform is adapted for full image searching to achieve better and faster matching results. Experimental results are given to demonstrate the improvements made by our proposed method.
Real-time freehand 3D ultrasound (US) has attracted much attention in clinical practices because it provides interactive feedback to help the clinicians acquire not only high-quality images but also timely information of the scanning area, which is necessary in intraoperative examinations. In this study, we developed a real-time freehand 3D US imaging system which can obtain volume reconstruction and visualisation during data acquisition at real-time level. Based on two popular algorithms, i.e. squared distance weighted interpolation (SDW) and Bezier interpolation, we designed corresponding parallel computing methods that were implemented on the graphics processing unit (GPU) to incrementally reconstruct and display the tissues being scanned using Visualisation toolkits (VTK). With a typical B-scan image size of 302 × 268 at an acquisition rate of 25 Hz and a preset volume size of 90 × 81 × 192, the system achieved an incremental reconstruction-visualisation rate of up to 32 frames/s and 119 frames/s for the SDW and Bezier algorithms, respectively, achieving the real-time 3D US.
The past decade has witnessed great advances in three-dimensional (3-D) medical ultrasound (US) imaging instrumentation. An increasing demand for portable 3-D US equipment is one of the main trends upcoming in the market. In this study, we developed a low cost, portable, sensorless and wireless 3-D US imaging system. A laptop US scanner with a conventional linear probe and a convex probe was used to acquire 2-D US B-scans. A client program was developed and run on the US scanner for capturing the pictures of screen during a freehand scanning without a positional sensor, and then the JPEG compression was applied to the pictures for reducing the image data size. The image data was sent to a remote workstation in real-time through Wi-Fi connection. A neural network model was used to recognize the characters (e.g. imaging depth and probe model information) displayed on the screen of the US scanner. The server on the remote workstation communicated with the US scanner, received raw image data, and finally reconstructed 3-D US images. The positions of the B-scans were obtained by estimating the spacings of B-scan image sequence, which was learned by measuring adaptive speckle decorrelation curves in mechanically collected B-scan frames. The performance of the proposed system has been demonstrated through experiments conducted on a US resolution phantom in vitro as well as human tissues in vivo.
Freehand 3-D ultrasound (US) produces 3-D volume data of anatomical objects from a sequence of irregularly located 2-D B-mode US images (B-scans). In 3-D US, the voxel intensities are calculated by interpolating those pixels from raw B-scans. Current interpolation algorithms do not consider sparsity of the raw data and are time consuming in computation. In this paper, we aim to perform the 3-D reconstruction of freehand US with sparse raw data in a more efficient manner. A novel interpolation algorithm takes advantage of Bezier curves. A single sweep of raw B-scans is collected, and the third-order Bezier curves are employed for approximating the voxels located in a control window. In in vitro and in vivo experiments, a fetus phantom and a subject's forearm were scanned using the freehand 3-D US system and reconstructed using the proposed Bezier interpolation algorithm and three popular interpolation algorithms, respectively. The results showed that the proposed algorithm significantly outperformed the other three algorithms when the raw B-scans were relatively sparse and the interpolation error in gray level can be reduced by 0.51-5.07. The speed for 3-D reconstruction can be improved by 90.6-97.2 because a single third-order Bezier curve using four control points (i.e., the pixel points) is able to estimate more than four voxels, whereas the estimation of a voxel value often requires a number of pixels in conventional techniques.
By the use of the way of real analysis, we estimate the weight functions and give
some new Hilbert-type integral inequalities in the whole plane with nonhomogeneous
kernels and multiparameters. The constant factors related to the hypergeometric function
and the beta function are proved to be the best possible. We also consider the
equivalent forms, the reverses, and some particular cases in the homogeneous kernels.
Speckle suppression plays an important role in improving ultrasound (US) image quality. While lots of algorithms have been proposed for 2D US image denoising with remarkable filtering quality, there is relatively less work done on 3D ultrasound speckle suppression, where the whole volume data rather than just one frame needs to be considered. Then, the most crucial problem with 3D US denoising is that the computational complexity increases tremendously. The nonlocal means (NLM) provides an effective method for speckle suppression in US images. In this paper, a programmable graphic-processor-unit- (GPU-) based fast NLM filter is proposed for 3D ultrasound speckle reduction. A Gamma distribution noise model, which is able to reliably capture image statistics for Log-compressed ultrasound images, was used for the 3D block-wise NLM filter on basis of Bayesian framework. The most significant aspect of our method was the adopting of powerful data-parallel computing capability of GPU to improve the overall efficiency. Experimental results demonstrate that the proposed method can enormously accelerate the algorithm.
Freehand three-dimensional ultrasound imaging is a highly attractive research area because it is capable of volumetric visualization and analysis of tissues and organs. The reconstruction algorithm plays a key role to the construction of three-dimensional ultrasound volume data with higher image quality and faster reconstruction speed. However, a systematic approach to such problem is still missing. A new fast marching method (FMM) for three-dimensional ultrasound volume reconstruction using the tracked and hand-held probe is proposed in this paper. Our reconstruction approach consists of two stages: bin-filling stage and hole-filling stage. Each pixel in the B-scan images is traversed and its intensity value is assigned to its nearest voxel in the bin-filling stage. For the efficient and accurate reconstruction, we present a new hole-filling algorithm based on the fast marching method. Our algorithm advances the interpolation boundary along its normal direction and fills the area closest to known voxel points in first, which ensure that the structural details of image can be preserved. Experimental results on both ultrasonic abdominal phantom and in vivo urinary bladder of human subject and comparisons with some popular algorithms are used to demonstrate its improvement in both reconstruction accuracy and efficiency.
A nonlocal quadratic functional of weighted differences is examined. The weights are based on image features and represent the affinity between different pixels in the image. By prescribing different formulas for the weights, one can generalize many local and nonlocal linear denoising algorithms, including the nonlocal means filter and the bilateral filter. In this framework one can easily show that continuous iterations of the generalized filter obey certain global characteristics and converge to a constant solution. The linear operator associated with the Euler-Lagrange equation of the functional is closely related to the graph Laplacian. We can thus interpret the steepest descent for minimizing the functional as a nonlocal diffusion process. This formulation allows a convenient framework for nonlocal variational minimizations, including variational denoising, Bregman iterations, and the recently proposed inverse scale space. It is also demonstrated how the steepest descent flow can be used for segmentation. Following kernel based methods in machine learning, the generalized diffusion process is used to propagate sporadic initial user’s information to the entire image. Unlike classical variational segmentation methods, the process is not explicitly based on a curve length energy and thus can cope well with highly nonconvex shapes and corners. Reasonable robustness to noise is still achieved.
Partial Differential equations (PDE), wavelets-based methods and neighborhood filters were proposed as locally adaptive machines
for noise removal. Recently, Buades, Coll and Morel proposed the Non-Local (NL-) means filter for image denoising. This method replaces a noisy pixel by the weighted average of other image pixels with weights reflecting
the similarity between local neighborhoods of the pixel being processed and the other pixels. The NL-means filter was proposed as an intuitive neighborhood filter but theoretical connections to diffusion and non-parametric estimation approaches are also given by the authors. In this
paper we propose another bridge, and show that the NL-means filter also emerges from the Bayesian approach with new arguments. Based on this observation, we show how the performance of this
filter can be significantly improved by introducing adaptive local dictionaries and a new statistical distance measure to
compare patches. The new Bayesian NL-means filter is better parametrized and the amount of smoothing is directly determined by the noise variance (estimated from image data)
given the patch size. Experimental results are given for real images with artificial Gaussian noise added, and for images
with real image-dependent noise.
3D freehand ultrasound imaging is becoming a widespread technique in medical examinations. This imaging technique produces a set of irregularly spaced B-scans. Reconstructing a regular grid from these B-scans is a challenging problem that enables the visualization and further analysis of the acquired data. This paper focuses on extending an existing method [1] to define the output reconstruction grid based on principal component analysis (PCA). Our method introduces a model for the region of interest (ROI) in order to adapt the grid to the ROI. In addition, a technique based on normalized convolution is proposed for the interpolation problem. A new applicability function based on the correlation function of a linear probe is used to avoid inter-resolution cell blurring.
We propose the use of nonlocal operators to define new types of flows and functionals for image processing and elsewhere. A main advantage over classical PDE-based algorithms is the ability to handle better textures and repetitive structures. This topic can be viewed as an extension of spectral graph theory and the diffusion geometry framework to functional analysis and PDE-like evolutions. Some possible applications and numerical examples are given, as is a general framework for approximating Hamilton-Jacobi equations on arbitrary grids in high demensions, e.g., for control theory.
Inspired by the recent work of Bertalmio, Sapiro, Caselles, and Ballester [Technical report, ECE-University of Minnesota (1999)], we intend to develop general mathematical models for local deterministic inpaintings. In smooth regions, inpaintings are connected to the harmonic and biharmonic extensions, and inpainting orders are dened and analyzed. For inpaintings involving the recovery of edges, we propose a variational model closely connected to the celebrated total variational (TV) denoising and deblurring model of Rudin, Osher, and Fatemi [Physica D, 60 (1992), pp. 259-268]. Such a model and its algorithm intrinsically combine the denoising and inpainting processes. We demonstrate its eectiveness through the applications in restoring scratched old photos, disocclusions in computer vision, and separating texts from images. 1 Introduction Perhaps the best way to explain inpainting is to quote the rst paragraph of the very recent paper of Bertalmio, Sapiro, Caselles and...
The class of ℓ 1 -regularized optimization problems has received much attention recently because of the introduction of “compressed sensing”, which allows images and signals to be reconstructed from small amounts of data. Despite this recent attention, many ℓ 1 -regularized problems still remain difficult to solve, or require techniques that are very problem-specific. In this paper, we show that Bregman iteration can be used to solve a wide variety of constrained optimization problems. Using this technique, we propose a “split Bregman” method, which can solve a very broad class of ℓ 1 -regularized problems. We apply this technique to the Rudin-Osher-Fatemi functional for image denoising and to a compressed sensing problem that arises in magnetic resonance imaging.
DOI: 10.1016/j.ultras.2004.05.003 3D ultrasound is a promising imaging modality for clinical diagnosis and treatment monitoring. Its cost is relatively low in comparison with CT and MRI, no intensive training and radiation protection is required for its operation, and its hardware is movable and can potentially be portable. In this study, we developed a portable freehand 3D ultrasound imaging system for the assessment of musculoskeletal body parts. A portable ultrasound scanner was used to obtain real-time B-mode ultrasound images of musculoskeletal tissues and an electromagnetic spatial sensor was fixed on the ultrasound probe to acquire the position and orientation of the images. The images were digitized with a video digitization device and displayed with its orientation and position synchronized in real-time with the data obtained by the spatial sensor. A program was developed for volume reconstruction, visualization, segmentation and measurement using Visual C++ and Visualization toolkits (VTK) software. A 2D Gaussian filter and a Median filter were implemented to improve the quality of the B-scan images collected by the portable ultrasound scanner. An improved distance-weighted grid-mapping algorithm was proposed for volume reconstruction. Temporal calibrations were conducted to correct the delay between the collections of images and spatial data. Spatial calibrations were performed using a cross-wire phantom. The system accuracy was validated by one cylinder and two cuboid phantoms made of silicone. The average errors for distance measurement in three orthogonal directions in comparison with micrometer measurement were 0.06 ± 0.39, −0.27 ± 0.27, and 0.33 ± 0.39 mm, respectively. The average error for volume measurement was −0.18% ± 5.44% for the three phantoms. The system has been successfully used to obtain the volume images of a fetus phantom, the fingers and forearms of human subjects. For a typical volume with 126 × 103 × 109 voxels, the 3D image could be reconstructed from 258 B-scans (640 × 480 pixels) within one minute using a portable PC with Pentium IV 2.4 GHz CPU and 512 MB memories. It is believed that such a portable volume imaging system will have many applications in the assessment of musculoskeletal tissues because of its easy accessibility. Author name used in this publication: Q. H. Huang Author name used in this publication: Y. P. Zheng Author name used in this publication: M. H. Lu Author name used in this publication: Z. R. Chi Centre for Signal Processing, Department of Electronic and Information Engineering Rehabilitation Engineering Centre
Visualized freehand 3-D ultrasound reconstruction offers to image incremental reconstruction during acquisition and guide users to scan interactively for high-quality volumes. We originally used the graphics processing unit (GPU) to develop a visualized reconstruction algorithm that achieves real-time level. Each newly acquired image was transferred to the memory of the GPU and inserted into the reconstruction volume on the GPU. The partially reconstructed volume was then rendered using GPU-based incremental ray casting. After visualized reconstruction, hole-filling was performed on the GPU to fill remaining empty voxels in the reconstruction volume. We examine the real-time nature of the algorithm using in vitro and in vivo datasets. The algorithm can image incremental reconstruction at speed of 26-58 frames/s and complete 3-D imaging in the acquisition time for the conventional freehand 3-D ultrasound.
A critical issue in image restoration is the problem of noise removal while keeping the integrity of relevant image information. Denoising is a crucial step to increase image quality and to improve the performance of all the tasks needed for quantitative imaging analysis. The method proposed in this paper is based on a 3-D optimized blockwise version of the nonlocal (NL)-means filter (Buades, et al., 2005). The NL-means filter uses the redundancy of information in the image under study to remove the noise. The performance of the NL-means filter has been already demonstrated for 2-D images, but reducing the computational burden is a critical aspect to extend the method to 3-D images. To overcome this problem, we propose improvements to reduce the computational complexity. These different improvements allow to drastically divide the computational time while preserving the performances of the NL-means filter. A fully automated and optimized version of the NL-means filter is then presented. Our contributions to the NL-means filter are: 1) an automatic tuning of the smoothing parameter; 2) a selection of the most relevant voxels; 3) a blockwise implementation; and 4) a parallelized computation. Quantitative validation was carried out on synthetic datasets generated with BrainWeb (Collins, et al., 1998). The results show that our optimized NL-means filter outperforms the classical implementation of the NL-means filter, as well as two other classical denoising methods [anisotropic diffusion (Perona and Malik, 1990)] and total variation minimization process (Rudin, et al., 1992) in terms of accuracy (measured by the peak signal-to-noise ratio) with low computation time. Finally, qualitative results on real data are presented .
The search for efficient image denoising methods is still a valid challenge at the crossing of functional analysis and statistics. In spite of the sophistication of the recently proposed methods, most algorithms have not yet attained a desirable level of applicability. All show an out- standing performance when the image model corresponds to the algorithm assumptions but fail in general and create artifacts or remove image fine structures. The main focus of this paper is, first, to define a general mathematical and experimental methodology to compare and classify classical image denoising algorithms and, second, to propose a nonlocal means (NL-means) algorithm ad- dressing the preservation of structure in a digital image. The mathematical analysis is based on the analysis of the “method noise,” defined as the difference between a digital image and its denoised version. The NL-means algorithm is proven to be asymptotically optimal under a generic statistical image model. The denoising performance of all considered methods are compared in four ways; mathematical: asymptotic order of magnitude of the method noise under regularity assumptions; perceptual-mathematical: the algorithms artifacts and their explanation as a violation of the image model; quantitative experimental: by tables of L2 distances of the denoised version to the original image. The most powerful evaluation method seems, however, to be the visualization of the method noise on natural images. The more this method noise looks like a real white noise, the better the method.
In image processing, restoration is expected to improve the qualitative inspection of the image and the performance of quantitative image analysis techniques. In this paper, an adaptation of the nonlocal (NL)-means filter is proposed for speckle reduction in ultrasound (US) images. Originally developed for additive white Gaussian noise, we propose to use a Bayesian framework to derive a NL-means filter adapted to a relevant ultrasound noise model. Quantitative results on synthetic data show the performances of the proposed method compared to well-established and state-of-the-art methods. Results on real images demonstrate that the proposed method is able to preserve accurately edges and structural details of the image.
This paper presents a novel approach for speckle reduction and coherence enhancement of ultrasound images based on nonlinear coherent diffusion (NCD) model. The proposed NCD model combines three different models. According to speckle extent and image anisotropy, the NCD model changes progressively from isotropic diffusion through anisotropic coherent diffusion to, finally, mean curvature motion. This structure maximally low-pass filters those parts of the image that correspond to fully developed speckle, while substantially preserving information associated with resolved-object structures. The proposed implementation algorithm utilizes an efficient discretization scheme that allows for real-time implementation on commercial systems. The theory and implementation of the new technique are presented and verified using phantom and clinical ultrasound images. In addition, the results from previous techniques are compared with the new method to demonstrate its performance.
3D freehand ultrasound imaging is a very attractive technique in medical examinations and intra-operative stage for its cost and field of view capacities. This technique produces a set of non parallel B-scans which are irregularly distributed in the space. Reconstruction amounts to computing a regular lattice volume and is needed to apply conventional computer vision algorithms like registration. In this paper, a new 3D reconstruction method is presented, taking explicitly into account the probe trajectory. Experiments were conducted on different data sets with various probe motion types and indicate that this technique outperforms classical methods, especially on low acquisition frame rate.
This paper provides the derivation of speckle reducing anisotropic diffusion (SRAD), a diffusion method tailored to ultrasonic and radar imaging applications. SRAD is the edge-sensitive diffusion for speckled images, in the same way that conventional anisotropic diffusion is the edge-sensitive diffusion for images corrupted with additive noise. We first show that the Lee and Frost filters can be cast as partial differential equations, and then we derive SRAD by allowing edge-sensitive anisotropic diffusion within this context. Just as the Lee and Frost filters utilize the coefficient of variation in adaptive filtering, SRAD exploits the instantaneous coefficient of variation, which is shown to be a function of the local gradient magnitude and Laplacian operators. We validate the new algorithm using both synthetic and real linear scan ultrasonic imagery of the carotid artery. We also demonstrate the algorithm performance with real SAR data. The performance measures obtained by means of computer simulation of carotid artery images are compared with three existing speckle reduction schemes. In the presence of speckle noise, speckle reducing anisotropic diffusion excels over the traditional speckle removal filters and over the conventional anisotropic diffusion method in terms of mean preservation, variance reduction, and edge localization.
The wide recognition that emerging nano-devices will be inherently unreliable motivates the evaluation of information processing algorithms running on noisy hardware as well as the design of robust schemes for reliable performance against hardware errors of varied characteristics. In this paper, we investigate the performance of a popular statistical inference algorithm, belief propagation (BP) on probabilistic graphical models, implemented on noisy hardware, and we propose two robust implementations of the BP algorithm targeting different computation noise distributions. We assume that the BP messages are subject to zero-mean transient additive computation noise. We focus on graphical models satisfying the contraction mapping condition that guarantees the convergence of the noise-free BP. We first upper bound the distances between the noisy BP messages and the fixed point of (noise-free) BP as a function of the iteration number. Next, we propose two implementations of BP, namely, censoring BP and averaging BP, that are robust to computation noise. Censoring BP rejects incorrect computations to keep the algorithm on the right track to convergence, while averaging BP takes the average of the messages in all iterations up to date to mitigate the effects of computation noise. Censoring BP works effectively when, with high probability, the computation noise is exactly zero, and averaging BP, although having a slightly larger overhead, works effectively for general zero-mean computation noise distributions. Sufficient conditions on the convergence of censoring BP and averaging BP are derived. Simulations on the Ising model demonstrate that the two proposed implementations successfully converge to the fixed point achieved by noise-free BP. Additionally, we apply averaging BP to a BP-based image denoising algorithm and as a BP decoder for LDPC codes. In the image denoising application, averaging BP successfully denoises an image even when nominal BP fails to do so in the - resence of computation noise. In the BP LDPC decoder application, the power of averaging BP is manifested by the reduction in the residual error rates compared with the nominal BP decoder.
We aimed to investigate the role of invariant natural killer T (iNKT) cells in infection-associated pregnancy loss. Wild-type (WT) C57BL/6 mice and iNKT cell-deficient Jα18(-/-) mice were treated with lipopolysaccharide (LPS). Embryo resorption rates (ERRs), decidual costimulatory molecule and activation molecule expression, and cytokine production were determined. WT and Jα18(-/-) mice were adoptively transferred with purified iNKT cells. ERRs, decidual costimulatory molecule and activation molecule expression, and cytokine production were assessed. LPS-treated Jα18(-/-) mice showed markedly reduced ERRs, decreased CD40, CD80, CD86, and CD69 expression, and reduced Th1 cytokine production at the maternal-fetal interface compared with WT mice. ERRs, expression of CD40, CD80, CD86, and CD69, and Th1 cytokine production in LPS-injected Jα18(-/-) mice following iNKT cell adoptive transfer were remarkably upregulated compared with control mice that did not receive adoptively transferred iNKT cells. Our results suggest that iNKT cells play an important role in LPS-induced pregnancy loss.
B-Mode ultrasound is often used during neurosurgery to provide intra-operative images of the brain though a craniotomy, but the use of D ultrasound during surgery is still in its infancy. We have developed a system that provides real-time freehand D ultrasound reconstruction at a reduced resolution. The reconstruction proceeds incrementally and the 3D image is overlayed, via a computer, on a pre-operative 3D MRI scan. This provides the operator with the necessary feedback to maintain a constant freehand sweep-rate, and also ensures that the sweep covers the desired anatomical volume. All of the ultrasound video frames are buffered, and a full-resolution, compounded reconstruction proceeds once the manual sweep is complete. We have also developed tools for manual tagging of homologous landmarks in the 3D MRI and D ultrasound volumes that use a piecewise cubic approximation of thin-plate spline interpolation to achieve interactive nonlinear registration and warping of the MRI volume to the ultrasound volume: Each time a homologous point-pair is identified by the use, the image of the warped MRI is updated on the computer screen after less than 0.5 s.
Three-dimensional ultrasound (3DUS) offers a valuable approach to many diagnostic imaging problems and provides a relatively low cost alternative to D imaging modalities such as MRI and CT. To obtain a 3DUS image, a clinical ultrasound scanner is used to produce a sequence of adjacent 'slices' of the region of interest, which are recorded and used to reconstruct a volume. The required location of each slice may be found either by a priori knowledge of the scan path (mechanized scanning) or by tracking the path of the transducer at scan time (freehand scanning). For some imaging applications the scan path is irregular, making mechanized scanning difficult or impossible. Some examples are: imaging vascular disease in the carotid arteries, joint imaging and obstetrics. We have developed a freehand 3DUS system that is capable of such imaging tasks. We used phantoms containing a wire and a 50% stenosed cylinder phantom to test our system for vascular applications. We found the mean error in locating any image plane to be less than 0.1 mm ((sigma) equals 0.05 mm), with minimal geometric distortion. The degree of stenosis of the phantom was measured from a reconstructed volume to be 52.6% ((sigma) equals 2%). We have started a clinical stenosis study and a pilot joint imaging study. Our clinical results indicate that measurement of stenosis in patients should be possible.
This article discusses the design features of an interactive system for acquiring, analyzing, and displaying volume sonographic patient data. Methods for reprojection of two-dimensional (2D) sonographic image data into a volume matrix are discussed. We describe an intuitive, easy-to-use graphical user interface that facilitates physician operation of a system incorporating an interactive volume renderer for optimization of viewing orientation and data presentation and incorporates stereoscopic viewing. Visualization methods are described that permit the operator interactively to extract tissues or organs of interest from the rest of the volume scanned. Selected examples of clinical images are given to demonstrate system capability. The system represents a cost-effective 3D ultrasound system integrating clinical scanners and graphics workstations. © 1997 John Wiley & Sons, Inc. Int J Imaging Syst Technol, 8: 26–37, 1997
Objective: This paper describes an algorithm which resulted in a practical method for estimating an enclosed volume using three-dimensional ultrasound. It was tested in vitro on thin-walled phantoms. Method: Data was acquired with a standard mechanical sector transducer which was tilted through a given angle (26°, 51° or 88°) by a motorized mechanical holder. A total of 81 frames was captured, stored digitally on a Unix workstation, and scanned and converted into a three-dimensional volumetric data set. Planar contours were drawn manually to indicate an enclosing volume. From the given set of contours, a polyhedron was reconstructed and the volume calculated. Four different principles of manual contour indication; outside, inside, center, and leading edge to leading edge, were investigated. The volume estimation method was tested on a condom filled with water with varying volume. Results: The experiments demonstrated lowest error for the center edge contour indication, mean difference −0.47 ml ± 2.15 ml (mean ± 2 S.D.) in the volume range 1.15–45.90 ml. An inter-observer error of 0.6% ± 5.0% was found using the center edge contour indication. The error for leading edge contour identification was similar, 0.78 ml ± 2.65 ml. Both inside and outside contour identification gave much larger errors. Conclusion: It was concluded that this volume estimation method was accurate, contributed to low inter-observer error, and that the results from the different tracing indications demonstrated the necessity to standardize these procedures for thin-walled organs.
In three-dimensional (3D) freehand ultrasound (US), reconstructing a set of B-scans into a regular voxel array is the key procedure for consequent visualization and analysis. This paper presents a new adaptive interpolation algorithm for computing the voxel array to suppress speckle noises and enhance contrast. The local statistics of homogeneous regions including mean and variance were measured and the ratio of variance to mean was used as homogeneity criteria. For the computation of each voxel, the interpolation method was adaptively determined with respect to its local statistics. If the neighbouring pixels of a voxel satisfied the homogeneity criterion, its value was computed with an arithmetic mean filter. Otherwise, the voxel was probably locating in an inhomogeneous region and an adaptive distance-weighted (ADW) interpolation method was employed to compute its value. A resolution phantom and a subject’s forearm were reconstructed using the proposed algorithm and two other well-known methods – conventional distance-weighted (DW) and voxel nearest neighbourhood (VNN) interpolations. The comparison results demonstrated that the adaptive interpolation algorithm was able to suppress speckles, preserve edges and enhance contrast effectively for the volume reconstruction.
In this paper, we present a new method for simple acquisition of dynamic three-dimensional (3-D) ultrasound data. We used a magnetic position sensor device attached to the ultrasound probe for spatial location of the probe, which was slowly tilted in the transthoracic scanning position. The 3-D data were recorded in 10–20 s, and the analysis was performed on an external PC within 2 min after transferring the raw digital ultrasound data directly from the scanner. The spatial and temporal resolutions of the reconstruction were evaluated, and were superior to video-based 3-D systems. Examples of volume reconstructions with better than 7 ms temporal resolution are given. The raw data with Doppler measurements were used to reconstruct both blood and tissue velocity volumes. The velocity estimates were available for optimal visualization and for quantitative analysis. The freehand data reconstruction accuracy was tested by volume estimation of balloon phantoms, giving high correlation with true volumes. Results show in vivo 3-D reconstruction and visualization of mitral and aortic valve morphology and blood flow, and myocardial tissue velocity. We conclude that it was possible to construct multimodality 3-D data in a limited region of the human heart within one respiration cycle, with reconstruction errors smaller than the resolution of the original ultrasound beam, and with a temporal resolution of up to 150 frames per second.
This paper describes an algorithm for the reconstruction of 3D medical objects from ultrasound images. Reconstruction is performed by filtering and interpolating the available data using an optimal Bayesian criterion. A Rayleigh model is adopted to describe the image formation process.
The most attractive feature of 2D B-mode ultrasound for intra-operative use is that it is both a real time and a highly interactive modality. Most 3D freehand reconstruction methods, however, are not fully interactive because they do not allow the display of any part of the D ultrasound image until all data collection and reconstruction is finished. We describe a technique whereby the D reconstruction occurs in real-time as the data is acquired, and where the operator can view the progress of the reconstruction on three orthogonal slice views through the ultrasound volume. Capture of the ultrasound data can be immedi- ately followed by a straightforward, interactive nonlinear registration of a pre-operative MRI volume to match the intra-operative ultrasound. We demonstrate the our system on a deformable, multi-modal PVA-cryogel phantom and during a clinical surgery.
This paper presents a new three-dimensional (3D) ultrasound reconstruction algorithm for generation of 3D images from a series of two-dimensional (2D) B-scans acquired in the mechanical linear scanning framework. Unlike most existing 3D ultrasound reconstruction algorithms, which have been developed and evaluated in the freehand scanning framework, the new algorithm has been designed to capitalize the regularity pattern of the mechanical linear scanning, where all the B-scan slices are precisely parallel and evenly spaced. The new reconstruction algorithm, referred to as the Cyclic Regularized Savitzky-Golay (CRSG) filter, is a new variant of the Savitzky-Golay (SG) smoothing filter. The CRSG filter has been improved upon the original SG filter in two respects: First, the cyclic indicator function has been incorporated into the least square cost function to enable the CRSG filter to approximate nonuniformly spaced data of the unobserved image intensities contained in unfilled voxels and reduce speckle noise of the observed image intensities contained in filled voxels. Second, the regularization function has been augmented to the least squares cost function as a mechanism to balance between the degree of speckle reduction and the degree of detail preservation. The CRSG filter has been evaluated and compared with the Voxel Nearest-Neighbor (VNN) interpolation post-processed by the Adaptive Speckle Reduction (ASR) filter, the VNN interpolation post-processed by the Adaptive Weighted Median (AWM) filter, the Distance-Weighted (DW) interpolation, and the Adaptive Distance-Weighted (ADW) interpolation, on reconstructing a synthetic 3D spherical image and a clinical 3D carotid artery bifurcation in the mechanical linear scanning framework. This preliminary evaluation indicates that the CRSG filter is more effective in both speckle reduction and geometric reconstruction of 3D ultrasound images than the other methods.
Conventional interpolation algorithms for reconstructing freehand three-dimensional (3D) ultrasound data always contain speckle noises and artifacts. This paper describes a new algorithm for reconstructing regular voxel arrays with reduced speckles and preserved edges. To study speckle statistics properties including mean and variance in sequential B-mode images in 3D space, experiments were conducted on an ultrasound resolution phantom and real human tissues. In the volume reconstruction, the homogeneity of the neighborhood for each voxel was evaluated according to the local variance/mean of neighboring pixels. If a voxel was locating in a homogeneous region, its neighboring pixels were averaged as the interpolation output. Otherwise, the size of the voxel neighborhood was contracted and the ratio was re-calculated. If its neighborhood was deemed as an inhomogeneous region, the voxel value was calculated using an adaptive Gaussian distance weighted method with respect to the local statistics. A novel method was proposed to reconstruct volume data set with economical usage of memory. Preliminary results obtained from the phantom and a subject's forearm demonstrated that the proposed algorithm was able to well suppress speckles and preserve edges in 3D images. We expect that this study can provide a useful imaging tool for clinical applications using 3D ultrasound.
This article describes a fully automatic, real-time, freehand ultrasound calibration system. The system was designed to be simple and sterilizable, intended for operating-room usage. The calibration system employed an automatic-error-retrieval and accuracy-control mechanism based on a set of ground-truth data. Extensive validations were conducted on a data set of 10,000 images in 50 independent calibration trials to thoroughly investigate the accuracy, robustness, and performance of the calibration system. On average, the calibration accuracy (measured in three-dimensional reconstruction error against a known ground truth) of all 50 trials was 0.66 mm. In addition, the calibration errors converged to submillimeter in 98% of all trials within 12.5 s on average. Overall, the calibration system was able to consistently, efficiently and robustly achieve high calibration accuracy with real-time performance. (E-mail: [email protected]
/* */).
Stereotactic ultrasonography is a technique for determining the position and orientation of B-mode ultrasound images in a reference coordinate system. A technique for constructing three-dimensional (3D) image volumes has been developed that uses this new technology. Given several registered images, a 3D volume is constructed either by a "nearest-neighbor" or a "closest-points" interpolation approach. The resulting volume can be rendered using 3D rendering software. In addition, the voxels in the volume are at known positions allowing determination of position for structures in the volume. Results are shown for various test cases, and applicability to medical imaging applications and stereotactic neurosurgery is discussed.
A system is described that rapidly produces a regular 3-dimensional (3-D) data block suitable for processing by conventional image analysis and volume measurement software. The system uses electromagnetic spatial location of 2-dimensional (2-D) freehand-scanned ultrasound B-mode images, custom-built signal-conditioning hardware, UNIX-based computer processing and an efficient 3-D reconstruction algorithm. Utilisation of images from multiple angles of insonation, "compounding," reduces speckle contrast, improves structure coherence within the reconstructed grey-scale image and enhances the ability to detect structure boundaries and to segment and quantify features. Volume measurements using a series of water-filled latex and cylindrical foam rubber phantoms with volumes down to 0.7 mL show that a high degree of accuracy, precision and reproducibility can be obtained. Extension of the technique to handle in vivo data sets by allowing physiological criteria to be taken into account in selecting the images used for construction is also illustrated.
The objective of this article is to provide scientists, engineers and clinicians with an up-to-date overview on the current state of development in the area of three-dimensional ultrasound (3-DUS) and to serve as a reference for individuals who wish to learn more about 3-DUS imaging. The sections will review the state of the art with respect to 3-DUS imaging, methods of data acquisition, analysis and display approaches. Clinical sections summarize patient research study results to date with discussion of applications by organ system. The basic algorithms and approaches to visualization of 3-D and 4-D ultrasound data are reviewed, including issues related to interactivity and user interfaces. The implications of recent developments for future ultrasound imaging/visualization systems are considered. Ultimately, an improved understanding of ultrasound data offered by 3-DUS may make it easier for primary care physicians to understand complex patient anatomy. Tertiary care physicians specializing in ultrasound can further enhance the quality of patient care by using high-speed networks to review volume ultrasound data at specialization centers. Access to volume data and expertise at specialization centers affords more sophisticated analysis and review, further augmenting patient diagnosis and treatment.
Three-dimensional freehand ultrasound imaging produces a set of irregularly spaced B-scans, which are typically reconstructed on a regular grid for visualization and data analysis. Most standard reconstruction algorithms are designed to minimize computational requirements and do not exploit the underlying shape of the data. We investigate whether an approximation with splines holds any promise as a better reconstruction method. A radial basis function approximation method is implemented and compared with three standard methods. While the radial basis approach is computationally expensive, it produces accurate reconstructions without the kind of visible artefacts common with the standard methods. The other potential advantages of radial basis functions, such as the direct computation of derivatives, make further investigation worthwhile.
Conventional freehand three-dimensional (3-D) ultrasound is a multi-stage process. First, the clinician scans the area of interest. Next, the ultrasound data is used to construct a 3-D voxel array, which can then be visualized by, for example, any-plane slicing. The strict separation of data acquisition and visualization disturbs the interactive nature of the ultrasound examination. Furthermore, some systems require the clinician to wait for an unacceptable amount of time while the voxel array is constructed. In this paper, we describe a novel freehand 3-D ultrasound system which allows accurate acquisition of the raw data and immediate visualization of arbitrary slices through the data. Minimal processing separates the acquisition and visualization processes: in particular, at no stage is a voxel array constructed. Instead, the standard graphics hardware found inside most desktop computers is exploited to synthesize arbitrary slices directly from the raw B-scans.
Although recent studies have demonstrated the potential value of compounded data for improvement in signal-to-noise ratio and speckle contrast for three-dimensional (3-D) ultrasonography, clinical applications are lacking. We investigated the potential of six degrees-of-freedom (6-DOF) scanhead position and orientation measurement (POM) devices for registration of in vivo multiplanar, irregularly sampled ultrasound (US) images to a regular 3-D volume space. The results demonstrate that accurate spatial and temporal registration of four-dimensional (4-D) US data can be achieved using a 6-DOF scanhead tracking system. For reconstruction of arbitrary, irregularly sampled US data, we introduce a technique based upon a weighted, ellipsoid Gaussian convolution kernel. Volume renderings of 3-D and 4-D compounded in vivo US data are presented. The results, although restricted to the field of cerebrovascular disease, will be of value to other applications of 3-D sonography, particularly those in which compounding of data through irregular sampling may provide superior information on tissue or vessel structure.
This paper presents a multiscale algorithm for the reconstruction of human anatomy from a set of ultrasound (US) images. Reconstruction is formulated in a Bayesian framework as an optimization problem with a large number of unknown variables. Human tissues are represented by the interpolation of coefficients associated to the nodes of a 3-D cubic grid. The convergence of the Bayesian method is usually slow and initialization dependent. In this paper, a multiscale approach is proposed to increase the convergence rate of the iterative process of volume estimation. A coarse estimate of the volume is first obtained using a cubic grid with a small number of nodes initialized with a constant value computed from the observed data. The volume estimate is then recursively improved by refining the grid step. Experimental results are provided to show that multiscale method achieves faster convergence rates compared with a single-scale approach. This is the key improvement toward real-time implementations. Experimental results of 3-D reconstruction of human anatomy are presented to assess the performance of the algorithm and comparisons with the single-scale method are presented.
Several techniques have been described in the literature in recent years for the reconstruction of a regular volume out of a series of ultrasound (US) slices with arbitrary orientations, typically scanned by means of US freehand systems. However, a systematic approach to such a problem is still missing. This paper focuses on proposing a theoretical framework for the 3-D US volume reconstruction problem. We introduce a statistical method for the construction and trimming of the sampling grid where the reconstruction will be carried out. The results using in vivo US data demonstrate that the computed reconstruction grid that encloses the region-of-interest (ROI) is smaller than those obtained from other reconstruction methods in those cases where the scanning trajectory deviates from a pure straight line. In addition, an adaptive Gaussian interpolation technique is studied and compared with well-known interpolation methods that have been applied to the reconstruction problem in the past. We find that the proposed method numerically outperforms former proposals in several control studies; subjective visual results also support this conclusion and highlight some potential deficiencies of methods previously proposed.
This paper describes a high-definition freehand 3-D ultrasound (US) system, with accuracy surpassing that of previously documented systems. 3-D point location accuracy within a US data set can be achieved to within 0.5 mm. Such accuracy is possible through a series of novel system-design and calibration techniques. The accuracy is quantified using a purpose-built tissue-mimicking phantom, designed to create realistic clinical conditions without compromising the accuracy of the measurement procedure. The paper includes a thorough discussion of the various ways of measuring system accuracy and their relative merits; and compares, in this context, all recently documented freehand 3-D US systems.
Volume reconstruction is a key procedure in 3D ultrasound imaging. An algorithm named as squared-distance-weighted (SDW) interpolation has been earlier proposed to reduce the blurring effect in the 3D ultrasonic images caused by the conventional distance weighted (DW) interpolation. However, the SDW parameter alpha, which controls the weight distribution, is a constant assigned by operators so that the interpolation effect is invariant for both sharp edges and speckle noises. In this paper, we introduced a new adaptive algorithm based on SDW interpolation for volume reconstruction of 3D freehand ultrasound. In the algorithm, the local statistics of pixels surrounding each voxel grid were used to adaptively adjust the parameter alpha in SDW. The voxel grids with a higher ratio of local variance and mean in their neighbourhoods would have a smaller alpha to make the image details sharper, while the voxel grids locating in regions with a lower ratio of local variance and mean would have a larger alpha to smooth image content in homogeneous regions, where speckle noise is usually observed and damages the image quality. By comparing the simulation results using the SDW and new adaptive algorithm, it was demonstrated that this new algorithm worked well in both edge preservation and speckle reduction.
Segmenting cardiac ultrasound images requires a model for the statistics of speckle in the images. Although the statistics of speckle are well understood for the raw transducer signal, the statistics of speckle in the image are not. This paper evaluates simple empirical models for first-order statistics for the distribution of gray levels in speckle. The models are created by analyzing over 100 images obtained from commercial ultrasound machines in clinical settings. The data in the images suggests a unimodal scalable family of distributions as a plausible model. Four families of distributions (Gamma, Weibull, Normal, and Log-normal) are compared with the data using goodness-of-fit and misclassification tests. Attention is devoted to the analysis of artifacts in images and to the choice of goodness-of-fit and misclassification tests. The distribution of parameters of one of the models is investigated and priors for the distribution are suggested.
This paper presents methods and a clinical procedure for integrating B-mode ultrasound images tagged with position information with a planning computed tomography (CT) scan for radiotherapy. A workflow is described that allows the integration of these modalities into the clinic. A surface mapping approach provides a preregistration of the ultrasound image borders onto the patient's skin. Successively, a set of individual ultrasound images from a freehand sweep is chosen by the physician. These images are automatically registered with the planning CT scan using novel intensity-based methods. We put a particular focus on deriving an appropriate similarity measure based on the physical properties and artifacts of ultrasound. A combination of a weighted mutual information term, edge correlation, clamping to the skin surface, and occlusion detection is able to assess the alignment of structures in ultrasound images and information reconstructed from the CT data. We demonstrate the practicality of our methods on five patients with head and neck tumors and cervical lymph node metastases and provide a detailed report on the conducted experiments, including the setup, calibration, acquisition, and verification of our algorithms. The mean target registration error on nine data sets is 3.9 mm. Thus, the additional information about intranodal architecture and fulfillment of malignancy criteria derived from a high-resolution ultrasonography of lymph nodes can be localized and visualized in the CT scan coordinate space and is made available for further radiation treatment planning.
This paper aims to apply median filters for reducing interpolation error and improving the quality of 3D images in a freehand 3D ultrasound (US) system. BACKGROUND AND MOTIVATION: Freehand 3D US imaging has been playing an important role in obtaining the entire 3D impression of tissues and organs. Reconstructing a sequence of irregularly located 2D US images (B-scans) into a 3D data set is one of the key procedures for visualization and data analysis.
In this study, we investigated the feasibility of using median filters for the reconstruction of 3D images in a freehand 3D US system. The B-scans were collected using a 7.5 MHz ultrasound probe. Four algorithms including the standard median (SM), Gaussian weighted median (GWM) and two types of distance-weighted median (DWM) filters were proposed to filter noises and compute voxel intensities. Qualitative and quantitative comparisons were made among the results of different methods based on the image set captured in freehand from the forearm of a healthy subject. A leave-one-out approach was used to demonstrate the performance of the median filters for predicting the removed B-scan pixels.
Compared with the voxel nearest-neighbourhood (VNN) and distance-weighted (DW) interpolation methods, the four median filters reduced the interpolation error by 8.0-24.0% and 1.2-21.8%, respectively, when 1/4 to 5 B-scans was removed from the raw B-scan sequence.
In summary, the median filters can improve the quality of volume reconstruction by reducing the interpolation errors and facilitate the following image analyses in clinical applications.
A method for reducing speckle noise in medical ultrasonic images
is presented. It is called the adaptive weighted median filter (AWMF)
and is based on the weighted median, which originates from the
well-known median filter through the introduction of weight
coefficients. By adjusting the weight coefficients and consequently the
smoothing characteristics of the filter according to the local
statistics around each point of the image, it is possible to suppress
noise while edges and other important features are preserved.
Application of the filter to several ultrasonic scans has shown that
processing improves the detectability of small structures and subtle
gray-scale variations without affecting the sharpness or anatomical
information of the original image. Comparison with the pure median
filter demonstrates the superiority of adaptive techniques over their
space-invariant counterparts. Examples of processed images show that the
AWMF preserves small details better than other nonlinear space-varying
filters which offer equal noise reduction in uniform areas
Previous extensions of two-dimensional ultrasonic imaging to three
dimensions used lattice diagrams which give measurement information, but
no anatomic detail. The authors conducted three sets of experiments to
test the hypothesis that complete acoustic backscatter data should be
retained to produce useful information about heart structure and
function. First, in vitro compound B-scans were taken under ideal
conditions; second, in vitro rotating conventional sector scans were
taken to test clinically applicable methods; and third, clinical in vivo
rotating conventional sector scans were taken of a human volunteer. It
is concluded that the resulting images show details of cardiac anatomy
and have great clinical promise. Interactive analysis and surface and
volume displays give context and perspective information which should
improve diagnostic accuracy, communication with noncardiologists and
yield more precise measurements of anatomical structure and function
Fast Marching Methods are numerical schemes for computing solutions to the non-linear Eikonal equation and related static Hamilton-Jacobi equations. Based on entropy-satisfying upwind schemes and fast sorting techniques, they yield consistent, accurate, and highly efficient algorithms. They are optimal in the sense that the computational complexity of the algorithms is O(N log N ), where N is the total number of points in the domain. The schemes are of use in a variety of applications, including problems in shape offsetting, computing distances from complex curves and surfaces, shape-from-shading, photolithographic development, computing rst arrivals in seismic travel times, construction of shortest geodesics on surfaces, optimal path planning around obstacles, and visibility and reection calculations. In this paper, we review the development of these techniques, including the theoretical and numerical underpinnings, provide details of the computational schemes including hig...