No full-text available
To read the full-text of this research,
you can request a copy directly from the author.
Cluster tracking algorithms for a digital image-based elasto-tomography system
... Manual Palpation is still currently an important method for breast cancer detection and regular self-examination is considered an important preventative measure. However, the size of the cancerous lump has to exceed 1-2 cm in order to be located easily [Hii, 2005] and by this stage the cancerous cells may have spread beyond the breast. Mammography represents the principal and most effective technology currently available for breast cancer screening. ...
Magnetic Resonance Elastography (MRE) is an emerging method for non-invasive breast cancer screening. It takes the MRI displacement data output and reconstructs the internal stiffness distribution, where cancerous tissue is approximately five to ten times stiffer than healthy breast tissue. Hence, MRE offers a high contrast solution to this diagnostic problem. Current MRE methods for reconstructing stiffness use forward simulation based optimization methods that are highly non-linear, non-convex and very heavy computationally. This research develops integral-based inverse problem solutions that reformulate the underlying differential equations in terms of integrals of MRI measured displacement data, and this transforms the problem into a linear, convex optimization. All derivative terms in the formulation are removed by special choice of integration limits, so no smoothing or filtering of the input data is required. The resulting equations can easily be solved by linear least squares requiring very minimal computation. 1D inverse algorithms were developed to provide a proof of concept of the integral-based method. Initially, the complete compressible 2D Navier's equations were used to develop the 2D inverse methods. Reasonable results were achieved with the algorithm successfully identifying a 1cm by 1cm tumour with up to 10% noise, data resolution of 20 measured points per cm and actuation frequencies of 100Hz. However, for the same input data set, a simplified incompressible 2D model was used as the basis for the final proposed inverse algorithm. This approach significantly improved results by removing ill-conditioned terms from the original formulation. For a 1cm by 1 cm tumour, accurate results were obtained with up to 40% noise, a range of actuation frequencies and very low data resolution of the order of 2 measured points per cm. These results thus indicate that more crude and less expensive data measurement systems could be used to obtain good results. The methods developed can be readily extended to 3D by applying a similar incompressible integral formulation to the 3D Navier's equations.
Two-dimensional discrete fourier transform (2D fft)
Scitation is the online home of leading journals and conference proceedings from AIP Publishing and AIP Member Societies
: Report developed under SBIR contract. A two-color particle imaging velocimetry system was demonstrated for obtaining two-dimensional velocity measurements in a subsonic reacting flow and in a supersonic flowfield. Data was obtained using three different camera systems, an 8-mm video camera, a digital color CCD camera system and a 35-mm camera. The data obtained from the wall boundary layer region in the supersonic wind tunnel showed good correlation with LDV data. The flow in the wake region of the flame holder was investigated using the PIV system developed. The results obtained through these experiments showed that the PIV system was capable of resolving the flow information in the highly turbulent region with a high degree of accuracy. Values of shear layer stresses behind the bluff body region were also calculated from the PIV data. Reacting flow experiments showed the ability of the PIV system to resolve velocity fields in a highly turbulent, large gradient flow regime. The software yielded RMS fluctuations in the 1.5% range which is less than the 3% limit widely reported in the literature. Taitech, Inc. is currently commercializing the PIV system developed in this contract.
In this paper, we present an algorithm for fast calculation of the normalized cross correlation (NCC) and its applica-tion to the problem of template matching. Given a template t, whose position is to be determined in an image f , t h e basic idea of the algorithm is to represent the template, for which the normalized cross correlation is calculated, as a sum of rectangular basis functions. Then the correlation is calculated for each basis function instead of the whole template. T h e r e s u l t o f the correlation of the template t and the image f is obtained as the weighted sum of the correlation functions of the basis functions. Depending on the approximation, the algorithm can by far outperform Fourier{transform based implementations of the normalized cross correlation algorithm and it is especially suited to problems, where many diierent templates are to be found in the same image f .
This paper presents various optimisation that can be applied to the sum of absolute differences (SAD) correlation algorithm for automated landmark detection. This has applications in mobile robotic navigation and mapping. We show how some assumptions about the environment and the generic form of strong landmarks selected by the SAD correlation algorithm have led to the development of an algorithm to enable near real tune selection of strong landmarks from visual information. The landmarks that have been selected from a series of frames using our optimisation are shown to be stable through the image sequence, demonstration the scale invariance of the landmarks that are selected by the SAD correlation algorithm.
This paper describes a visual object detection framework that is capable of processing images extremely rapidly while achieving high detection rates. There are three key contributions. The first is the introduction of a new image representation called the "Integral Image" which allows the features used by our detector to be computed very quickly. The second is a learning algorithm, based on AdaBoost, which selects a small number of critical visual features and yields extremely efficient classifiers (6). The third contribution is a method for combining classifiers in a "cascade" which allows background regions of the image to be quickly discarded while spending more computation on promising object-like regions. A set of experiments in the domain of face detection are presented. The system yields face detection performace comparable to the best previous systems (18, 13, 16, 12, 1). Implemented on a conventional desktop, face detection proceeds at 15 frames per second.
Image registration finds a variety of applications in computer vision. Unfortunately, traditional image registration techniques tend to be costly. We present a new image registration technique that makes use of the spatial intensity gradient of the images to find a good match using a type of Newton-Raphson iteration. Our technique is taster because it examines far fewer potential matches between the images than existing techniques Furthermore, this registration technique can be generalized to handle rotation, scaling and shearing. We show how our technique can be adapted tor use in a stereo vision system.
To evaluate the dynamic range of tissue imaged by elastography, the mechanical behavior of breast and prostate tissue samples subject to compression loading has been investigated. A model for the loading was validated and used to guide the experimental design for data collection. The model allowed the use of small samples that could be considered homogeneous; this assumption was confirmed by histological analysis. The samples were tested at three strain rates to evaluate the viscoelastic nature of the material and determine the validity of modeling the tissue as an elastic material for the strain rates of interest. For loading frequencies above 1 Hz, the storage modulus accounted for over 93 percent of the complex modulus. The data show that breast fat tissue has a constant modulus over the strain range tested while the other tissues have a modulus that is dependent on the strain level. The fibrous tissue samples from the breast were found to be 1 to 2 orders of magnitude stiffer than fat tissue. Normal glandular breast tissue was found to have an elastic modulus similar to that of fat at low strain levels, but the modulus of the glandular tissue increased by an order of magnitude above fat at high strain levels. Carcinomas from the breast were stiffer than the other tissues at the higher strain level; intraductal in situ carcinomas were like fat at the low strain level and much stiffer than glandular tissue at the high strain level. Infiltrating ductal carcinomas were much stiffer than any of the other breast tissues. Normal prostate tissue has a modulus that is lower than the modulus of the prostate cancers tested. Tissue from prostate with benign prostatic hyperplasia (BPH) had modulus values significantly lower than normal tissue. There was a constant but not significant difference in the modulus of tissues taken from the anterior and posterior portions of the gland.
This paper describes an inverse reconstruction technique based on a modified Newton Raphson iterative scheme and the finite element method, which has been developed for computing the spatial distribution of Young's modulus from within soft tissues. Computer simulations were conducted to determine the relative merits of reconstructing tissue elasticity using knowledge of (a) known displacement boundary conditions (DBC), and (b) known stress boundary conditions (SBC). The results demonstrated that computing Young's modulus using knowledge of SBC allows accurate quantification of Young's modulus. However, the quality of the images produced using this reconstruction approach was dependent on the Young's modulus distribution assumed at the start of the reconstruction procedure. Computing Young's modulus from known DBC provided relative estimates of tissue elasticity which, despite the disadvantage of not being able to accurately quantify Young's modulus, formed images that were generally superior in quality to those produced using the known SBC, and were not affected by the trial solution.
The results of preliminary experiments on phantoms demonstrated that this reconstruction technique is capable in practice of improving the fidelity of tissue elasticity images, reducing the artefacts otherwise present in strain images, and recovering Young's modulus images that possess excellent spatial and contrast resolution.
Changes in vessel wall elasticity may be indicative of vessel pathologies. It is known, for example, that the presence of plaque stiffens the vascular wall, and that the heterogeneity of its composition may lead to plaque rupture and thrombosis. Another domain of application where ultrasound elastography may be of interest is the study of vascular wall elasticity to predict the risk of aneurysmal tissue rupture. In this paper, this technology is introduced as an approach to noninvasively characterize superficial arteries. In such a case, a linear array ultrasound transducer is applied on the skin over the region of interest, and the arterial tissue is dilated by the normal cardiac pulsation. The elastograms, the equivalent elasticity images, are computed from the assessment of the vascular tissue motion. Investigating the forward problem, it is shown that motion parameters might be difficult to interpret; that is because tissue motion occurs radially within the vessel wall while the ultrasound beam propagates axially. As a consequence of that, the elastograms are subjected to hardening and softening artefacts, which are to be counteracted. In this paper, the Von Mises (VM) coefficient is proposed as a new parameter to circumvent such mechanical artefacts and to appropriately characterize the vessel wall. Regarding the motion assessment, the Lagrangian estimator was used; that is because it provides the full two-dimensional strain tensor necessary to compute the VM coefficient. The theoretical model was validated with biomechanical simulations of the vascular wall properties. The results allow believing in the potential of the method to differentiate hard plaques and lipid pools from normal vascular tissue. Potential in vivo implementation of noninvasive vascular elastography to characterize abdominal aneurysms and superficial arteries such as the femoral and the carotid is discussed.
A new Kalman-filter based active contour model is proposed for
tracking of nonrigid objects in combined spatio-velocity space. The
model employs measurements of gradient-based image potential and of
optical-flow along the contour as system measurements. In order to
improve robustness to image clutter and to occlusions an optical-flow
based detection mechanism is proposed. The method detects and rejects
spurious measurements which are not consistent with previous estimation
of image motion
Although it is well known that cross correlation can be efficiently implemented in the transform domain, the normalized form of cross correlation preferred for feature matching applications does not have a simple frequency domain expression. Normalized cross correlation has been computed in the spatial domain for this reason. This short paper shows that unnormalized cross correlation can be efficiently normalized using precomputing integrals of the image and image 2 over the search window. 1
Image preprocessing seeks to modify and prepare the pixel values of a digitised image to produce a form that is more suitable for subsequent operations within the generic model. There are two major branches of image preprocessing, namely image enhancement and image restoration.
Many researchers have proposed imaging the stiffness distribution in breast tissue to enhance diagnosis of disease. They suppose that cancers are much stiffer than the surrounding tissue but to our knowledge no measurements have been made of these properties that accurately characterize them over a wide range of strain. We hypothesize that there is a correlation between elastic modulus in compression and histological diagnosis (e.g. infiltrating ductal carcinoma, normal glandular tissue, etc.). We also hypothesize that the cancer exhibits greater non-linearity; its change in modulus with strain is greater. We present a correlation that allows elastic moduli to be estimated from force displacement curves measured during punch indentation testing. The tissue samples tested were obtained during surgery and were tested immediately after removal from the body. We found that there is a significant difference in the stiffness and the rate of increase in stiffness with strain between cancerous and benign breast tissues. Infiltrating ductal cancer is more than 10 times as stiff as normal fat tissue at 1% strain, and more than 70 times as stiff at 15% strain. Compared to normal glandular tissue, this type of cancer is more than 2.5 times as stiff at 1% strain and nearly 5 times as stiff at 15% strain. Therefore, relative stiffness is a good indicator of histological diagnosis.
Normalized cross correlation has been used extensively for many machine vision applications, but the traditional normalized correlation operation does not meet speed requirements for time-critical applications. In this paper, we propose a fast normalized cross correlation computation for defect detection application. A sum-table scheme is utilized, which allows the calculations of image mean, image variance and cross-correlation between images to be invariant to the size of template window. Given larger images of size N M × and the neighborhood window of size n m × , the computational complexity can be significantly reduced from) (N M n m O ⋅ ⋅ ⋅ with the traditional normalized correlation operation to only) (N M O ⋅ with the proposed sum-table scheme.
It has previously been noted that, for conventional machine code, there is a strong relationship be-tween static and dynamic code measurements. One of the goals of this paper is to examine whether this same relationship is true of Java programs at the bytecode level. To this end, the hypothesis of a linear correlation between static and dynamic frequencies was investigated using Pearson's correlation coefficient. Programs from the Java Grande and SPEC benchmarks suites were used in the analysis.
The normalized cross correlation (NCC) has been used extensively in machine vision for industrial inspection, but the traditional NCC suffers from false alarms for a complicated image that contains partial uniform regions. In this paper, we study the use of NCCs for defect detection in complicated images. The performance of NCCs in monochrome and color images, and the effect of image smoothing are empirically evaluated. The proposed NCC in a smoothed color image can effectively alleviate false alarms in defect detection applications.
We present a new robust point matching algorithm (RPM) that can
jointly estimate the correspondence and non-rigid transformations
between two point-sets that may be of different sizes. The algorithm
utilizes the soft assign for the correspondence and the thin-plate
spline for the non-rigid mapping. Embedded within a deterministic
annealing framework, the algorithm can automatically reject a fraction
of the points as outliers. Experiments on both 2D synthetic point-sets
with varying degrees of deformation, noise and outliers, and on real 3D
sulcal point-sets (extracted from brain MRI) demonstrate the robustness
of the algorithm
Template matching by normalized correlations is a common technique for determine the existence and compute the location of
a shape within an image. In many cases the run time of computer vision applications is dominated by repeated computation of
template matching, applied to locate multiple templates in varying scale and orientation. A straightforward implementation
of template matching for an image size n and a template size k requires order of kn operations. There are fast algorithms that require order of n log n operations. We describe a new approximation scheme that requires order n operations. It is based on the idea of “Integral-Images”, recently introduced by Viola and Jones.
A snake is an energy-minimizing spline guided by external constraint forces and influenced by image forces that pull it toward features such as lines and edges. Snakes are active contour models: they lock onto nearby edges, localizing them accurately. Scale-space continuation can be used to enlarge the capture region surrounding a feature. Snakes provide a unified account of a number of visual problems, including detection of edges, lines, and subjective contours; motion tracking; and stereo matching. We have used snakes successfully for interactive interpretation, in which user-imposed constraint forces guide the snake near features of interest.
Typescript (photocopy). "Date of submission March 2005." Thesis (M.E.M.)--University of Canterbury, 2005. Includes bibliographical references.
Digital Image-based Elasto-Tomography (DIET) is a novel method of determining the distribution of elastic properties within the breast. Using an array of calibrated digital cameras and in inverse reconstruction algorithm, DIET allows reconstruction of the internal elastic stiffness distribution of the breast using only motions at the breast surface. This reconstructed stiffness should clearly show carcinoma based on their high elastic property contrast with healthy tissue. Proof of concept studies are presented for both the callibration of the digital imaging system and the inverse reconstruction algorithm. The reconstruction algorithm identified high stiffness tumors in the majority of test cases, even with the addition of random noise based on expected calibration accuracy.
Tracking the speckle patterns produced by moving targets has been shown effective for angle independent imaging of blood flow and tissue motion. While speckle tracking overcomes major limitations of Doppler-based flow imaging, the computational complexity of commonly used cross correlation algorithms currently limits it to off-line studies. A much simpler algorithm for angle independent motion imaging is described in this paper. This method requires only one absolute difference operation per pixel, compared to eight operations for normalized cross correlation. Quantitative studies using speckle-generating targets translated by fixed amounts both axially and laterally indicate that the technique tracks moving speckle as accurately as correlation. Color flow images generated from clinical blood and liver data highlight the success of the technique for tracking both large and small motions in two dimensions. The algorithm's suitability for implementation in digital hardware makes possible the development of clinical instruments for angle independent ultrasonic imaging of blood flow and tissue motion in real time.
A nuclear magnetic resonance imaging (MRI) method is presented for quantitatively mapping the physical response of a material
to harmonic mechanical excitation. The resulting images allow calculation of regional mechanical properties. Measurements
of shear modulus obtained with the MRI technique in gel materials correlate with independent measurements of static shear
modulus. The results indicate that displacement patterns corresponding to cyclic displacements smaller than 200 nanometers
can be measured. The findings suggest the feasibility of a medical imaging technique for delineating elasticity and other
mechanical properties of tissue.
MR elastography is a novel imaging technique for the visualization of elastic properties of tissue. It is expected that this method will have diagnostic value for the clarification of suspicious breast lesions. Low-frequency mechanical waves are coupled into the tissue and visualized via an MR sequence which is phase-locked to the mechanical excitation. Commonly, elasticity is assumed to be isotropic and reconstruction is performed in only two dimensions. The technique is extended to three dimensions such that the entire symmetric elasticity tensor is assessed. This is achieved by measuring different phases of the mechanical wave during one oscillatory cycle. Thereby it is possible to provide information about the anisotropy of the elasticity tensor. Finite-element simulations as well as phantom experiments are performed to demonstrate the feasibility of the method. Initial clinical results of a breast carcinoma are presented. The analysis of the eigenvalues of the elasticity tensor support the hypothesis that breast carcinoma might exhibit an anisotropic elasticity distribution. The surrounding benign tissue appears isotropic. Thereby new and additional diagnostic information is provided which might help in distinguishing between benign and malignant breast diseases.
Over the past decade, several methods have been proposed to image tissue elasticity based on imaging methods collectively called elastography. While progress in developing these systems has been rapid, the basic understanding of tissue properties to interpret elastography images is generally lacking. To address this limitation, we developed a system to measure the Young's modulus of small soft tissue specimens. This system was designed to accommodate biological soft tissue constraints such as sample size, geometry imperfection and heterogeneity. The measurement technique consists of indenting an unconfined small block of tissue while measuring the resulting force. We show that the measured force–displacement slope of such a geometry can be transformed to the tissue Young's modulus via a conversion factor related to the sample's geometry and boundary conditions using finite element analysis. We also demonstrate another measurement technique for tissue elasticity based on quasi-static magnetic resonance elastography in which a tissue specimen encased in a gelatine–agarose block undergoes cyclical compression with resulting displacements measured using a phase contrast MRI technique. The tissue Young's modulus is then reconstructed from the measured displacements using an inversion technique. Finally, preliminary elasticity measurement results of various breast tissues are presented and discussed.
For more information on this article, see medicalphysicsweb.org
This study investigated the feasibility of using low-dose multidetector dynamic computed tomography (CT) scan for imaging breast. We measured the radiation dose using a phantom at low- and standard-dose CT. To compare the image quality at low- and standard-dose CT, we evaluated normal breasts in 57 cases. In 44 cases with breast cancer, we assessed the staging and time-enhancement curves of breast cancer. In conclusion, the low-dose multidetector dynamic CT scan is feasible for the evaluation of the breast, with reduced radiation dose and with similar image quality when compared with standard-dose CT scan. In breast cancers, low-dose dynamic CT could be used for the staging of breast cancer before surgery.
Elasticity is an important physical property of material. In the clinical practice, elasticity is used for physical examination in several ways, such as palpation or percussion. Differences in elasticity can help facilitate the diagnosis of tumors and their extent. Elasticity is an essential property in the diagnosis of liver cirrhosis, or soft degeneration in tissue necrosis. In addition, information of tissue elasticity is utilized in virtual reality systems such as telepalpation and computer assisted surgery. It was difficult to obtain such properties in vivo by using conventional measurement methods. To overcome this problem, magnetic resonance elastography (MRE) has been developed that provides noninvasive in vivo measurements of elasticity for human tissue. We summarize this MRE method in this paper. When an object is oscillated from the surface in a known frequency, acoustic strain waves propagate into the material and one can calculate the physical constants of a material elasticity by the wave velocity. In MRE measurements, a cyclic micromotion caused by the acoustic strain waves is obtained as an MR image that is synchronized to the oscillation. By measuring the local wavelength of the strain waves, we can obtain the elasticity constants. Several examples of MRE image including in vivo measurements are provided as well as several methods to estimate the local wavelength from MRE images are described.
Breast Cancer in Older Women
Fentiman 1.S, "Breast Cancer in Older Women," Breast Cancer Online,
vol. 5, 2002.
Magnetic Resonance Elastography by Direct Visualization of Propagating Acoustic Strain Waves
- R Muthupillai
- P J Rossman
- D J Lomas
- J F Greenleaf
- S J Riederer
- R L Ehman
R. Muthupillai, P. J. Rossman, D. J. Lomas, J. F. Greenleaf, S. J. Riederer,
and R. L. Ehman, "Magnetic Resonance Elastography by Direct
Visualization of Propagating Acoustic Strain Waves," Science, pp.
269:1854-1857,1995.
Tissue Actuation Studies for a Digital~Image Elasto-Tomography (DIET) Breast Cancer Screening System
- S Kinoshita
S. Kinoshita, "Tissue Actuation Studies for a Digital~Image Elasto-Tomography (DIET) Breast Cancer Screening System," in Mechanical
Engineering. Christchurch: Univ. of Canterbury, Master thesis, 2005.
Simple Linear Regression and Correlation Analysis
- M L Berenson
- D M Levine
- M Goldstein
M. L. Berenson, D. M. Levine, and M. Goldstein, "Simple Linear
Regression and Correlation Analysis," in Intermediate Statistical
Methods and Applications. New Jersey: Prentice-Hall, 1983, pp. 203-236.
- K Enslein
- A Ralston
- H S Wilf
K. Enslein, A. Ralston, and H. S. Wilf, Statistical Method for Digital
Computers, vol. 3. New York: Wiley-Interscience Publication, 1977.
Template Matching and Correlation
- J C Russ
J. C. Russ, "Template Matching and Correlation," in The Image
Processing Handbook, 2 ed. Raleigh: CRC Press, 1994, pp. 34]-346.