Detection of lines and boundaries in speckle images--application to medical ultrasound

Supervised segmentation for guiding peripheral revascularization with forward-viewing, robotically steered ultrasound guidewire

BACKGROUND

Approximately 500 000 patients present with critical limb ischemia (CLI) each year in the U.S., requiring revascularization to avoid amputation. While peripheral arteries can be revascularized via minimally invasive procedures, 25% of cases with chronic total occlusions are unsuccessful due to inability to route the guidewire beyond the proximal occlusion. Improvements to guidewire navigation would lead to limb salvage in a greater number of patients.

PURPOSE

Integrating ultrasound imaging into the guidewire could enable direct visualization of routes for guidewire advancement. In order to navigate a robotically-steerable guidewire with integrated imaging beyond a chronic occlusion proximal to the symptomatic lesion for revascularization, acquired ultrasound images must be segmented to visualize the path for guidewire advancement.

METHODS

The first approach for automated segmentation of viable paths through occlusions in peripheral arteries is demonstrated in simulations and experimentally-acquired data with a forward-viewing, robotically-steered guidewire imaging system. B-mode ultrasound images formed via synthetic aperture focusing (SAF) were segmented using a supervised approach (U-net architecture). A total of 2500 simulated images were used to train the classifier to distinguish the vessel wall and occlusion from viable paths for guidewire advancement. First, the size of the synthetic aperture resulting in the highest classification performance was determined in simulations (90 test images) and compared with traditional classifiers (global thresholding, local adaptive thresholding, and hierarchical classification). Next, classification performance as a function of the diameter of the remaining lumen (0.5 to 1.5 mm) in the partially-occluded artery was tested using both simulated (60 test images at each of 7 diameters) and experimental data sets. Experimental test data sets were acquired in four 3D-printed phantoms from human anatomy and six ex vivo porcine arteries. Accuracy of classifying the path through the artery was evaluated using microcomputed tomography of phantoms and ex vivo arteries as a ground truth for comparison.

RESULTS

An aperture size of 3.8 mm resulted in the best-performing classification based on sensitivity and Jaccard index, with a significant increase in Jaccard index (p < 0.05) as aperture diameter increased. In comparing the performance of the supervised classifier and traditional classification strategies with simulated test data, sensitivity and F1 score for U-net were 0.95 ± 0.02 and 0.96 ± 0.01, respectively, compared to 0.83 ± 0.03 and 0.41 ± 0.13 for the best-performing conventional approach, hierarchical classification. In simulated test images, sensitivity (p < 0.05) and Jaccard index both increased with increasing artery diameter (p < 0.05). Classification of images acquired in artery phantoms with remaining lumen diameters ≥ 0.75 mm resulted in accuracies > 90%, while mean accuracy decreased to 82% when artery diameter decreased to 0.5 mm. For testing in ex vivo arteries, average binary accuracy, F1 score, Jaccard index, and sensitivity each exceeded 0.9.

CONCLUSIONS

Segmentation of ultrasound images of partially-occluded peripheral arteries acquired with a forward-viewing, robotically-steered guidewire system was demonstrated for the first-time using representation learning. This could represent a fast, accurate approach for guiding peripheral revascularization.

Point detection in textured ultrasound images

Detection of point scatterers in textured ultrasound images can be challenging. This paper investigates how four multilook methods can improve the detection. We analyze many images with known point scatterer locations and randomly textured backgrounds. The normalized matched filter (NMF) and multilook coherence factor (MLCF) methods are normalized methods that do not require any texture correction prior to detection analysis. They are especially propitious when optimal texture correction of the ultrasound images is difficult to obtain. The results show significant improvement in detection performance when the MLCF method is weighted with the prewhitened and texture corrected image. The method can be applied even when we do not have prior knowledge about the optimal prewhitening limits. The multilook methods NMF and NMF weighted (NMFW) are very favorable methods to apply on images where acoustic noise dominates the speckle background.

Point Detection in Ultrasound Using Prewhitening and Multilook Optimization

We investigate methods to improve the detection of point scatterers in ultrasound imaging using the standard delay-and-sum (DAS) image as our starting point. An optimized whitening transform can increase the spatial resolution of the image. By splitting an image's frequency spectrum into many subsets using the multilook technique, we can exploit the coherent properties of a point scatterer. We present three new multilook methods and evaluate their effect on point detection. The performances are compared to DAS using synthetic aperture Field II simulations of a point scatterer in uniform speckle background. The results show that optimized prewhitening of the images can significantly improve the point detection. The multilook methods have the potential to improve the detection performance further when a sufficient number of looks are used. If prior knowledge about the optimal spectrum limits is unavailable and a nonoptimal prewhitening is applied, applying that the new multilook methods can considerably improve the point detection.

BUSIS: A Benchmark for Breast Ultrasound Image Segmentation

Breast ultrasound (BUS) image segmentation is challenging and critical for BUS computer-aided diagnosis (CAD) systems. Many BUS segmentation approaches have been studied in the last two decades, but the performances of most approaches have been assessed using relatively small private datasets with different quantitative metrics, which results in a discrepancy in performance comparison. Therefore, there is a pressing need for building a benchmark to compare existing methods using a public dataset objectively, to determine the performance of the best breast tumor segmentation algorithm available today, and to investigate what segmentation strategies are valuable in clinical practice and theoretical study. In this work, a benchmark for B-mode breast ultrasound image segmentation is presented. In the benchmark, (1) we collected 562 breast ultrasound images and proposed standardized procedures to obtain accurate annotations using four radiologists; (2) we extensively compared the performance of 16 state-of-the-art segmentation methods and demonstrated that most deep learning-based approaches achieved high dice similarity coefficient values (DSC ≥ 0.90) and outperformed conventional approaches; (3) we proposed the losses-based approach to evaluate the sensitivity of semi-automatic segmentation to user interactions; and (4) the successful segmentation strategies and possible future improvements were discussed in details.

Accurate model-based segmentation of gynecologic brachytherapy catheter collections in MRI-images

The gynecological cancer mortality rate, including cervical, ovarian, vaginal and vulvar cancers, is more than 20,000 annually in the US alone. In many countries, including the US, external-beam radiotherapy followed by high dose rate brachytherapy is the standard-of-care. The superior ability of MR to visualize soft tissue has led to an increase in its usage in planning and delivering brachytherapy treatment. A technical challenge associated with the use of MRI imaging for brachytherapy, in contrast to that of CT imaging, is the visualization of catheters that are used to place radiation sources into cancerous tissue. We describe here a precise, accurate method for achieving catheter segmentation and visualization. The algorithm, with the assistance of manually provided tip locations, performs segmentation using image-features, and is guided by a catheter-specific, estimated mechanical model. A final quality control step removes outliers or conflicting catheter trajectories. The mean Hausdorff error on a 54 patient, 760 catheter reference database was 1.49  mm; 51 of the outliers deviated more than two catheter widths (3.4  mm) from the gold standard, corresponding to catheter identification accuracy of 93% in a Syed-Neblett template. In a multi-user simulation experiment for evaluating RMS precision by simulating varying manually-provided superior tip positions, 3σ maximum errors were 2.44  mm. The average segmentation time for a single catheter was 3 s on a standard PC. The segmentation time, accuracy and precision, are promising indicators of the value of this method for clinical translation of MR-guidance in gynecologic brachytherapy and other catheter-based interventional procedures.

Line Detection as an Inverse Problem: Application to Lung Ultrasound Imaging

This paper presents a novel method for line restoration in speckle images. We address this as a sparse estimation problem using both convex and non-convex optimization techniques based on the Radon transform and sparsity regularization. This breaks into subproblems, which are solved using the alternating direction method of multipliers, thereby achieving line detection and deconvolution simultaneously. We include an additional deblurring step in the Radon domain via a total variation blind deconvolution to enhance line visualization and to improve line recognition. We evaluate our approach on a real clinical application: the identification of B-lines in lung ultrasound images. Thus, an automatic B-line identification method is proposed, using a simple local maxima technique in the Radon transform domain, associated with known clinical definitions of line artefacts. Using all initially detected lines as a starting point, our approach then differentiates between B-lines and other lines of no clinical significance, including Z-lines and A-lines. We evaluated our techniques using as ground truth lines identified visually by clinical experts. The proposed approach achieves the best B-line detection performance as measured by the F score when a non-convex [Formula: see text] regularization is employed for both line detection and deconvolution. The F scores as well as the receiver operating characteristic (ROC) curves show that the proposed approach outperforms the state-of-the-art methods with improvements in B-line detection performance of 54%, 40%, and 33% for [Formula: see text], [Formula: see text], and [Formula: see text], respectively, and of 24% based on ROC curve evaluations.

Development of a Low-Cost Medical Ultrasound Scanner Using a Monostatic Synthetic Aperture

OBJECTIVE

In this paper, we present the design of low-cost medical ultrasound scanners aimed at the detection of maternal mortality risk factors in developing countries.

METHOD

Modern ultrasound scanners typically employ a high element count transducer array with multichannel transmit and receive electronics. To minimize hardware costs, we employ a single piezoelectric element, mechanically swept across the target scene, and a highly cost-engineered single channel acquisition circuit. Given this constraint, we compare the achievable image quality of a monostatic fixed focus scanner (MFFS) with a monostatic synthetic aperture scanner (MSAS) using postfocusing. Quantitative analysis of image quality was carried out using simulation and phantom experiments, which were used to compare a proof-of-concept MSAS prototype with an MFFS device currently available on the market. Finally, in vivo experiments were performed to validate the MSAS prototype in obstetric imaging.

RESULTS

Simulations show that the achievable lateral resolution of the MSAS approach is superior at all ranges compared to the fixed focus approach. Phantom experiments verify the improved resolution of the MSAS prototype but reveal a lower signal to noise ratio. In vivo experiments show promising results using the MSAS for clinical diagnostics in prenatal care.

CONCLUSION

The proposed MSAS achieves superior resolution but lower SNR compared to an MFFS approach, principally due to lower acoustic energy emitted.

SIGNIFICANCE

The production costs of the proposed MSAS could be an order of magnitude lower than any other ultrasound system on the market today, bringing affordable obstetric imaging a step closer for developing countries.

An Artificial Immune System-Based Support Vector Machine Approach for Classifying Ultrasound Breast Tumor Images

A rapid and highly accurate diagnostic tool for distinguishing benign tumors from malignant ones is required owing to the high incidence of breast cancer. Although various computer-aided diagnosis (CAD) systems have been developed to interpret ultrasound images of breast tumors, feature selection and the setting of parameters are still essential to classification accuracy and the minimization of computational complexity. This work develops a highly accurate CAD system that is based on a support vector machine (SVM) and the artificial immune system (AIS) algorithm for evaluating breast tumors. Experiments demonstrate that the accuracy of the proposed CAD system for classifying breast tumors is 96.67%. The sensitivity, specificity, PPV, and NPV of the proposed CAD system are 96.67, 96.67, 95.60, and 97.48%, respectively. The receiver operator characteristic (ROC) area index A z is 0.9827. Hence, the proposed CAD system can reduce the number of biopsies and yield useful results that assist physicians in diagnosing breast tumors.

Hole filling with oriented sticks in ultrasound volume reconstruction

Volumes reconstructed from tracked planar ultrasound images often contain regions where no information was recorded. Existing interpolation methods introduce image artifacts and tend to be slow in filling large missing regions. Our goal was to develop a computationally efficient method that fills missing regions while adequately preserving image features. We use directional sticks to interpolate between pairs of known opposing voxels in nearby images. We tested our method on 30 volumetric ultrasound scans acquired from human subjects, and compared its performance to that of other published hole-filling methods. Reconstruction accuracy, fidelity, and time were improved compared with other methods.

Breast ultrasound despeckling using anisotropic diffusion guided by texture descriptors

Breast ultrasound (BUS) is considered the most important adjunct method to mammography for diagnosing cancer. However, this image modality suffers from an intrinsic artifact called speckle noise, which degrades spatial and contrast resolution and obscures the screened anatomy. Hence, it is necessary to reduce speckle artifacts before performing image analysis by means of computer-aided diagnosis systems, for example. In addition, the trade-off between smoothing level and preservation of lesion contour details should be addressed by speckle reduction schemes. In this scenario, we propose a BUS despeckling method based on anisotropic diffusion guided by Log-Gabor filters (ADLG). Because we assume that different breast tissues have distinct textures, in our approach we perform a multichannel decomposition of the BUS image using Log-Gabor filters. Next, the conduction coefficient of anisotropic diffusion filtering is computed using texture responses instead of intensity values as stated originally. The proposed algorithm is validated using both synthetic and real breast data sets, with 900 and 50 images, respectively. The performance measures are compared with four existing speckle reduction schemes based on anisotropic diffusion: conventional anisotropic diffusion filtering (CADF), speckle-reducing anisotropic diffusion (SRAD), texture-oriented anisotropic diffusion (TOAD), and interference-based speckle filtering followed by anisotropic diffusion (ISFAD). The validity metrics are the Pratt's figure of merit, for synthetic images, and the mean radial distance (in pixels), for real sonographies. Figure of merit and mean radial distance indices should tend toward '1' and '0', respectively, to indicate adequate edge preservation. The results suggest that ADLG outperforms the four speckle removal filters compared with respect to simulated and real BUS images. For each method--ADLG, CADF, SRAD, TOAD and ISFAD--the figure of merit median values are 0.83, 0.40, 0.39, 0.51 and 0.59, and the mean radial distance median results are 4.19, 6.29, 6.39, 6.43 and 5.88.

Stellate masses and histologic grades in breast cancer

Breast masses with a radiologic stellate pattern often transform into malignancies, but their tendency to be of low histologic grade yields a better survival rate compared with tumors with other patterns on mammography screening. This study was designed to investigate the correlation of histologic grade with stellate features extracted from the coronal plane of 3-D ultrasound images. A pre-processing method was proposed to facilitate the extraction of stellate features. Extracted features were statistically measured to derive a set of indices that quantitatively represent the stellate pattern. These indices then went through a selection procedure to build proper decision trees. The splitting rules of decision trees indicated that stellate tumors are associated with low grade. A set of indices from the low grade-associated rules has the potential to represent the stellate feature. Further investigation of the hypoechoic region of peripheral tissue is essential to establishment of a complete discriminating model for tumor grades.

Computer-aided multiview tumor detection for automated whole breast ultrasound

Automated whole breast ultrasound (ABUS) has become a popular screening tool in recent years. To reduce the review time and misdetection from ABUS images by physicians, a computer-aided detection (CADe) system for ABUS images based on a multiview method is proposed in this study. A total of 58 pathology-proven lesions from 41 patients were used to evaluate the performance of the system. In the proposed CADe system, the fuzzy c-mean clustering method was applied to detect tumor candidates from these ABUS images. Subsequently, the tumor likelihoods of these candidates could be estimated by a logistic linear regression model based on the intensity, morphology, location, and size features in the transverse, longitudinal, and coronal views. Finally, the multiview tumor likelihoods of the tumor candidates could be obtained from the estimated tumor likelihoods of the three views, and the tumor candidates with high multiview tumor likelihoods were regarded as the detected tumors in the proposed system. The sensitivities of the multiview tumor detection for selecting 5, 10, 20, and 30 tumor candidates with the largest multiview tumor likelihoods were 79.31%, 86.21%, 96.55%, and 98.28%, respectively.

An effective approach of lesion segmentation within the breast ultrasound image based on the cellular automata principle

In this paper, a novel lesion segmentation within breast ultrasound (BUS) image based on the cellular automata principle is proposed. Its energy transition function is formulated based on global image information difference and local image information difference using different energy transfer strategies. First, an energy decrease strategy is used for modeling the spatial relation information of pixels. For modeling global image information difference, a seed information comparison function is developed using an energy preserve strategy. Then, a texture information comparison function is proposed for considering local image difference in different regions, which is helpful for handling blurry boundaries. Moreover, two neighborhood systems (von Neumann and Moore neighborhood systems) are integrated as the evolution environment, and a similarity-based criterion is used for suppressing noise and reducing computation complexity. The proposed method was applied to 205 clinical BUS images for studying its characteristic and functionality, and several overlapping area error metrics and statistical evaluation methods are utilized for evaluating its performance. The experimental results demonstrate that the proposed method can handle BUS images with blurry boundaries and low contrast well and can segment breast lesions accurately and effectively.

Improved Rotating Kernel Transformation Based Contourlet Domain Image Denoising Framework

A contourlet domain image denoising framework based on a novel Improved Rotating Kernel Transformation is proposed, where the difference of subbands in contourlet domain is taken into account. In detail: (1). A novel Improved Rotating Kernel Transformation (IRKT) is proposed to calculate the direction statistic of the image; The validity of the IRKT is verified by the corresponding extracted edge information comparing with the state-of-the-art edge detection algorithm. (2). The direction statistic represents the difference between subbands and is introduced to the threshold function based contourlet domain denoising approaches in the form of weights to get the novel framework. The proposed framework is utilized to improve the contourlet soft-thresholding (CTSoft) and contourlet bivariate-thresholding (CTB) algorithms. The denoising results on the conventional testing images and the Optical Coherence Tomography (OCT) medical images show that the proposed methods improve the existing contourlet based thresholding denoising algorithm, especially for the medical images.

Whole breast lesion detection using naive bayes classifier for portable ultrasound

In recent years, portable PC-based ultrasound (US) imaging systems developed by some companies can provide an integrated computer environment for computer-aided diagnosis and detection applications. In this article, an automatic whole breast lesion detection system based on the naive Bayes classifier using the PC-based US system Terason t3000 (Terason Ultrasound, Burlington, MA, USA) with a hand-held probe is proposed. To easily retrieve the US images for any regions of the breast, a clock-based storing system is proposed to record the scanned US images. A computer-aided detection (CAD) system is also developed to save the physicians' time for a huge volume of scanned US images. The pixel classification of the US is based on the naive Bayes classifier for the proposed lesion detection system. The pixels of the US are classified into two types: lesions or normal tissues. The connected component labeling is applied to find the suspected lesions in the image. Consequently, the labeled two-dimensional suspected regions are separated into two clusters and further checked by two-phase lesion selection criteria for the determination of the real lesion, while reducing the false-positive rate. The free-response operative characteristics (FROC) curve is used to evaluate the detection performance of the proposed system. According to the experimental results of 31 cases with 33 lesions, the proposed system yields a 93.4% (31/33) sensitivity at 4.22 false positives (FPs) per hundred slices. Moreover, the speed for the proposed detection scheme achieves 12.3 frames per second (fps) with an Intel Dual-Core Quad 3 GHz processor and can be also effectively and efficiently used for other screening systems.

The 2D analytic signal for envelope detection and feature extraction on ultrasound images

The fundamental property of the analytic signal is the split of identity, meaning the separation of qualitative and quantitative information in form of the local phase and the local amplitude, respectively. Especially the structural representation, independent of brightness and contrast, of the local phase is interesting for numerous image processing tasks. Recently, the extension of the analytic signal from 1D to 2D, covering also intrinsic 2D structures, was proposed. We show the advantages of this improved concept on ultrasound RF and B-mode images. Precisely, we use the 2D analytic signal for the envelope detection of RF data. This leads to advantages for the extraction of the information-bearing signal from the modulated carrier wave. We illustrate this, first, by visual assessment of the images, and second, by performing goodness-of-fit tests to a Nakagami distribution, indicating a clear improvement of statistical properties. The evaluation is performed for multiple window sizes and parameter estimation techniques. Finally, we show that the 2D analytic signal allows for an improved estimation of local features on B-mode images.

Bio-magnetic signatures of fetal breathing movement

The purpose of fetal magnetoencephalography (fMEG) is to record and analyze fetal brain activity. Unavoidably, these recordings consist of a complex mixture of bio-magnetic signals from both mother and fetus. The acquired data include biological signals that are related to maternal and fetal heart function as well as fetal gross body and breathing movements. Since fetal breathing generates a significant source of bio-magnetic interference during these recordings, the goal of this study was to identify and quantify the signatures pertaining to fetal breathing movements (FBM). The fMEG signals were captured using superconducting quantum interference devices (SQUIDs) The existence of FBM was verified and recorded concurrently by an ultrasound-based video technique. This simultaneous recording is challenging since SQUIDs are extremely sensitive to magnetic signals and highly susceptible to interference from electronic equipment. For each recording, an ultrasound-FBM (UFBM) signal was extracted by tracing the displacement of the boundary defined by the fetal thorax frame by frame. The start of each FBM was identified by using the peak points of the UFBM signal. The bio-magnetic signals associated with FBM were obtained by averaging the bio-magnetic signals time locked to the FBMs. The results showed the existence of a distinctive sinusoidal signal pattern of FBM in fMEG data.

A novel approach to track fetal movement using multi-sensor magnetocardiographic recordings

Changes in fetal magnetocardiographic (fMCG) signals are indicators for fetal body movement. We propose a novel approach to reliably extract fetal body movements based on the field strength of the fMCG signal independent of its frequency. After attenuating the maternal MCG, we use a Hilbert transform approach to identify the R-wave. At each R-wave, we compute the center-of-gravity (cog) of the coordinate positions of MCG sensors, each weighted by the magnitude of the R-wave amplitude recorded at the corresponding sensor. We then define actogram as the distance between the cog computed at each R-wave and the average of the cog from all the R-waves in a 3-min duration. By applying a linear de-trending approach to the actogram we identify the fetal body movement and compare this with the synchronous occurrence of the acceleration in the fetal heart rate. Finally, we apply this approach to the fMCG recorded simultaneously with ultrasound from a single subject and show its improved performance over the QRS-amplitude based approach in the visually verified movements. This technique could be applied to transform the detection of fetal body movement into an objective measure of fetal health and enhance the predictive value of prevalent clinical testing for fetal wellbeing.

Coronary angiography enhancement for visualization

High quality visualization on X-ray angiograms is of great significance both for the diagnosis of vessel abnormalities and for coronary interventions. Algorithms for improving the visualization of detailed vascular structures without significantly increasing image noise are currently demanded in the market. A new algorithm called stick-guided lateral inhibition (SGLI) is presented for increasing the visibility of coronary vascular structures. A validation study was set up to compare the SGLI algorithm with the conventional unsharp masking (UM) algorithm on 20 still frames of coronary angiographic images. Ten experienced QCA analysts and nine cardiologists from various centers participated in the validation. Sample scoring value (SSV) and observer agreement value (OAV) were defined to evaluate the validation result, in terms of enhancing performance and observer agreement, respectively. The mean of SSV was concluded to be 77.1 ± 11.9%, indicating that the SGLI algorithm performed significantly better than the UM algorithm (P-value < 0.001). The mean of the OAV was concluded to be 70.3%, indicating that the average agreement with respect to a senior cardiologist was 70.3%. In conclusion, this validation study clearly demonstrates the superiority of the SGLI algorithm in the visualization of coronary arteries from X-ray angiograms.

Three Dimensional OCT in the Engineering of Tissue Constructs: A Potentially Powerful Tool for Assessing Optimal Scaffold Structure

Optical Coherence Tomography (OCT) provides detailed, real-time information on the structure and composition of constructs used in tissue engineering. The focus of this work is the OCT three-dimensional assessment of scaffolding architecture and distribution of cells on it. PLGA scaffolds were imaged in two and three-dimensions, both seeded and unseeded with cells. Then two types of scaffolds were reconstructed in three dimensions. Both scaffolding types were examined at three different seeding densities. The importance of three-dimensional assessments was evident, particularly with respect to porosity and identification of asymmetrical cell distribution.

Automated 2D-3D registration of portal images and CT data using line-segment enhancement

In prostate radiotherapy, setup errors with respect to the patient's bony anatomy can be reduced by aligning 2D megavoltage (MV) portal images acquired during treatment to a reference 3D kilovoltage (kV) CT acquired for treatment planning purposes. The purpose of this study was to evaluate a fully automated 2D-3D registration algorithm to quantify setup errors in 3D through the alignment of line-enhanced portal images and digitally reconstructed radiographs computed from the CT. The line-enhanced images were obtained by correlating the images with a filter bank of short line segments, or "sticks" at different orientations. The proposed methods were validated on (1) accurately collected gold-standard data consisting of a 3D kV cone-beam CT scan of an anthropomorphic phantom of the pelvis and 2D MV portal images in the anterior-posterior (AP) view acquired at 15 different poses and (2) a conventional 3D kV CT scan and weekly 2D MV AP portal images of a patient over 8 weeks. The mean (and standard deviation) of the absolute registration error for rotations around the right-lateral (RL), inferior-superior (IS), and posterior-anterior (PA) axes were 0.212 degree (0.214 degree), 0.055 degree (0.033 degree) and 0.041 degree (0.039 degree), respectively. The corresponding registration errors for translations along the RL, IS, and PA axes were 0.161 (0.131) mm, 0.096 (0.033) mm, and 0.612 (0.485) mm. The mean (and standard deviation) of the total registration error was 0.778 (0.543) mm. Registration on the patient images was successful in all eight cases as determined visually. The results indicate that it is feasible to automatically enhance features in MV portal images of the pelvis for use within a completely automated 2D-3D registration framework for the accurate determination of patient setup errors. They also indicate that it is feasible to estimate all six transformation parameters from a 3D CT of the pelvis and a single portal image in the AP view.

Level set segmentation of bovine corpora lutea in ex situ ovarian ultrasound images

BACKGROUND

The objective of this study was to investigate the viability of level set image segmentation methods for the detection of corpora lutea (corpus luteum, CL) boundaries in ultrasonographic ovarian images. It was hypothesized that bovine CL boundaries could be located within 1-2 mm by a level set image segmentation methodology.

METHODS

Level set methods embed a 2D contour in a 3D surface and evolve that surface over time according to an image-dependent speed function. A speed function suitable for segmentation of CL's in ovarian ultrasound images was developed. An initial contour was manually placed and contour evolution was allowed to proceed until the rate of change of the area was sufficiently small. The method was tested on ovarian ultrasonographic images (n = 8) obtained ex situ. A expert in ovarian ultrasound interpretation delineated CL boundaries manually to serve as a "ground truth". Accuracy of the level set segmentation algorithm was determined by comparing semi-automatically determined contours with ground truth contours using the mean absolute difference (MAD), root mean squared difference (RMSD), Hausdorff distance (HD), sensitivity, and specificity metrics.

RESULTS AND DISCUSSION

The mean MAD was 0.87 mm (sigma = 0.36 mm), RMSD was 1.1 mm (sigma = 0.47 mm), and HD was 3.4 mm (sigma = 2.0 mm) indicating that, on average, boundaries were accurate within 1-2 mm, however, deviations in excess of 3 mm from the ground truth were observed indicating under- or over-expansion of the contour. Mean sensitivity and specificity were 0.814 (sigma = 0.171) and 0.990 (sigma = 0.00786), respectively, indicating that CLs were consistently undersegmented but rarely did the contour interior include pixels that were judged by the human expert not to be part of the CL. It was observed that in localities where gradient magnitudes within the CL were strong due to high contrast speckle, contour expansion stopped too early.

CONCLUSION

The hypothesis that level set segmentation can be accurate to within 1-2 mm on average was supported, although there can be some greater deviation. The method was robust to boundary leakage as evidenced by the high specificity. It was concluded that the technique is promising and that a suitable data set of human ovarian images should be obtained to conduct further studies.

Quantification of thyroid volume using 3-D ultrasound imaging

Ultrasound (US) is among the most popular diagnostic techniques today. It is non-invasive, fast, comparably cheap, and does not require ionizing radiation. US is commonly used to examine the size, and structure of the thyroid gland. In clinical routine, thyroid imaging is usually performed by means of 2-D US. Conventional approaches for measuring the volume of the thyroid gland or its nodules may therefore be inaccurate due to the lack of 3-D information. This work reports a semi-automatic segmentation approach for the classification, and analysis of the thyroid gland based on 3-D US data. The images are scanned in 3-D, pre-processed, and segmented. Several pre-processing methods, and an extension of a commonly used geodesic active contour level set formulation are discussed in detail. The results obtained by this approach are compared to manual interactive segmentations by a medical expert in five representative patients. Our work proposes a novel framework for the volumetric quantification of thyroid gland lobes, which may also be expanded to other parenchymatous organs.

Pixel compounding: resolution-enhanced ultrasound imaging for quantitative analysis

Accurate measurement of structural features represented in medical images is important in clinical trials and patient diagnosis. A key factor for precision is spatial resolution, which in ultrasonic imaging is limited by transducer array arrangements, transmitting frequency, and data acquisition firmware. In this paper, a variation of pixel compounding is proposed to enhance ultrasound resolution using acquired cine loops. The technique operates on a sequence of ultrasound B-scan images acquired with random motion. Subpixel registration is estimated and a maximum a posteriori (MAP) approach with the shift information is used to reconstruct a high-resolution single image. A nonhomogeneous anisotropic diffusion algorithm follows from the estimation process and is implemented to enhance the high-resolution edges. Preliminary tests using simulations and phantom studies show promising results. Pixel compounding can be a powerful preprocessing tool to assure accurate segmentation, measurement, and analysis of ultrasound images.

Tunneling descent level set segmentation of ultrasound images

The presence of speckle in ultrasound images causes many spurious local minima in the energy function of active contours. These minima trap the segmentation prematurely under gradient descent and cause the algorithm to fail. This paper presents a substantially new reformulation of Tunneling Descent, which is a deterministic technique to escape from unwanted local minima. In the new formulation, the evolving curve is represented by level sets, and the evolution strategy is obtained as a sequence of constrained minimizations. The algorithm is used to segment the endocardium in 115 short axis cardiac ultrasound images. All segmentations are achieved without tweaking the energy function or numerical parameters. Experimental evaluation of the results shows that the algorithm overcomes multiple local minima to give segmentations that are considerably more accurate than conventional techniques.

Nonlinear diffusion in Laplacian pyramid domain for ultrasonic speckle reduction

A new speckle reduction method, i.e., Laplacian pyramid-based nonlinear diffusion (LPND), is proposed for medical ultrasound imaging. With this method, speckle is removed by nonlinear diffusion filtering of bandpass ultrasound images in Laplacian pyramid domain. For nonlinear diffusion in each pyramid layer, a gradient threshold is automatically determined by a variation of median absolute deviation (MAD) estimator. The performance of the proposed LPND method has been compared with that of other speckle reduction methods, including the recently proposed speckle reducing anisotropic diffusion (SRAD) and nonlinear coherent diffusion (NCD). In simulation and phantom studies, an average gain of 1.55 dB and 1.34 dB in contrast-to-noise ratio was obtained compared to SRAD and NCD, respectively. The visual comparison of despeckled in vivo ultrasound images from liver and carotid artery shows that the proposed LPND method could effectively preserve edges and detailed structures while thoroughly suppressing speckle. These preliminary results indicate that the proposed speckle reduction method could improve image quality and the visibility of small structures and fine details in medical ultrasound imaging.

Enhancing ultrasound texture differences for developing an in vivo 'virtual histology' approach to bovine ovarian imaging

A ‘virtual histology’ can be thought of as the ‘staining’ of a digital ultrasound image via image processing techniques in order to enhance the visualisation of differences in the echotexture of different types of tissues. Several candidate image-processing algorithms for virtual histology using ultrasound images of the bovine ovary were studied. The candidate algorithms were evaluated qualitatively for the ability to enhance the visual differences in intra-ovarian structures and quantitatively, using standard texture description features, for the ability to increase statistical differences in the echotexture of different ovarian tissues. Certain algorithms were found to create textures that were representative of ovarian micro-anatomical structures that one would observe in actual histology. Quantitative analysis using standard texture description features showed that our algorithms increased the statistical differences in the echotexture of stroma regions and corpus luteum regions. This work represents a first step toward both a general algorithm for the virtual histology of ultrasound images and understanding dynamic changes in form and function of the ovary at the microscopic level in a safe, repeatable and non-invasive way.

Semiautomatic 3-D prostate segmentation from TRUS images using spherical harmonics

Prostate brachytherapy quality assessment procedure should be performed while the patient is still on the operating table since this would enable physicians to implant additional seeds immediately into the prostate if necessary thus reducing the costs and increasing patient outcome. Seed placement procedure is readily performed under fluoroscopy and ultrasound guidance. Therefore, it has been proposed that seed locations be reconstructed from fluoroscopic images and prostate boundaries be identified in ultrasound images to perform dosimetry in the operating room. However, there is a key hurdle that needs to be overcome to perform the ultrasound and fluoroscopy-based dosimetry: it is highly time-consuming for physicians to outline prostate boundaries in ultrasound images manually, and there is no method that enables physicians to identify three-dimensional (3-D) prostate boundaries in postimplant ultrasound images in a fast and robust fashion. In this paper, we propose a new method where the segmentation is defined in an optimization framework as fitting the best surface to the underlying images under shape constraints. To derive these constraints, we modeled the shape of the prostate using spherical harmonics of degree eight and performed statistical analysis on the shape parameters. After user initialization, our algorithm identifies the prostate boundaries on the average in 2 min. For algorithm validation, we collected 30 postimplant prostate volume sets, each consisting of axial transrectal ultrasound images acquired at 1-mm increments. For each volume set, three experts outlined the prostate boundaries first manually and then using our algorithm. By treating the average of manual boundaries as the ground truth, we computed the segmentation error. The overall mean absolute distance error was 1.26 +/- 0.41 mm while the percent volume overlap was 83.5 +/- 4.2. We found the segmentation error to be slightly less than the clinically-observed interobserver variability.

Boundary Enhancement and Speckle Reduction for Ultrasound Images via Salient Structure Extraction

In this paper, we present an approach for medical ultrasound (US) image enhancement. It is based on a novel perceptual saliency measure which favors smooth, long curves with constant curvature. The perceptual salient boundaries of tissues in US images are enhanced by computing the saliency of directional vectors in the image space, via a local searching algorithm. Our measure is generally determined by curvature changes, intensity gradient and the interaction of neighboring vectors. To restrain speckle noise during the enhancement process, an adaptive speckle suspension term is also combined into the proposed saliency measure. The results obtained on both simulated images and medical US data reveal superior performance of the novel approach over a number of commonly used speckle filters. Applications of US image segmentation show that although the proposed algorithm cannot remove the speckle noise completely and may discard weak anatomical structures in some case, it still provides a considerable gain to US image processing for computer-aided diagnosis.

Ultrasound image segmentation: a survey

This paper reviews ultrasound segmentation paper methods, in a broad sense, focusing on techniques developed for medical B-mode ultrasound images. First, we present a review of articles by clinical application to highlight the approaches that have been investigated and degree of validation that has been done in different clinical domains. Then, we present a classification of methodology in terms of use of prior information. We conclude by selecting ten papers which have presented original ideas that have demonstrated particular clinical usefulness or potential specific to the ultrasound segmentation problem.

Automated 2D-3D registration of a radiograph and a cone beam CT using line-segment enhancement

The objective of this study was to develop a fully automated two-dimensional (2D)-three-dimensional (3D) registration framework to quantify setup deviations in prostate radiation therapy from cone beam CT (CBCT) data and a single AP radiograph. A kilovoltage CBCT image and kilovoltage AP radiograph of an anthropomorphic phantom of the pelvis were acquired at 14 accurately known positions. The shifts in the phantom position were subsequently estimated by registering digitally reconstructed radiographs (DRRs) from the 3D CBCT scan to the AP radiographs through the correlation of enhanced linear image features mainly representing bony ridges. Linear features were enhanced by filtering the images with "sticks," short line segments which are varied in orientation to achieve the maximum projection value at every pixel in the image. The mean (and standard deviations) of the absolute errors in estimating translations along the three orthogonal axes in millimeters were 0.134 (0.096) AP(out-of-plane), 0.021 (0.023) ML and 0.020 (0.020) SI. The corresponding errors for rotations in degrees were 0.011 (0.009) AP, 0.029 (0.016) ML (out-of-plane), and 0.030 (0.028) SI (out-of-plane). Preliminary results with megavoltage patient data have also been reported. The results suggest that it may be possible to enhance anatomic features that are common to DRRs from a CBCT image and a single AP radiography of the pelvis for use in a completely automated and accurate 2D-3D registration framework for setup verification in prostate radiotherapy. This technique is theoretically applicable to other rigid bony structures such as the cranial vault or skull base and piecewise rigid structures such as the spine.

2-D ultrasound strain images for breast cancer diagnosis using nonrigid subregion registration

Tissue elasticity of a lesion is a useful criterion for the diagnosis of breast ultrasound (US). Elastograms are created by comparing ultrasonic radio-frequency waveforms before and after a light-tissue compression. In this study, we evaluate the accuracy of continuous US strain image in the classification of benign from malignant breast tumors. A series of B-mode US images is applied and each case involves 60 continuous images obtained by using the steady artificial pressure of the US probe. In general, after compression by the US probe, a soft benign tumor will become flatter than a stiffened malignant tumor. We proposed a computer-aided diagnostic (CAD) system by utilizing the nonrigid image registration modality on the analysis of tumor deformation. Furthermore, we used some image preprocessing methods, which included the level set segmentation, to improve the performance. One-hundred pathology-proven cases, including 60 benign breast tumors and 40 malignant tumors, were used in the experiments to test the classification accuracy of the proposed method. Four characteristic values--normalized slope of metric value (NSM), normalized area difference (NAD), normalized standard deviation (NSD) and normalized center translation (NCT)--were computed for all cases. By using the support vector machine, the accuracy, sensitivity, specificity and positive and negative predictive values of the classification of continuous US strain images were satisfactory. The A(z) value of the support vector machine based on the four characteristic values used for the classification of solid breast tumors was 0.9358.

Segmentation of elastographic images using a coarse-to-fine active contour model

Delineation of radiofrequency-ablation-induced coagulation (thermal lesion) boundaries is an important clinical problem that is not well addressed by conventional imaging modalities. Elastography, which produces images of the local strain after small, externally applied compressions, can be used for visualization of thermal coagulations. This paper presents an automated segmentation approach for thermal coagulations on 3-D elastographic data to obtain both area and volume information rapidly. The approach consists of a coarse-to-fine method for active contour initialization and a gradient vector flow, active contour model for deformable contour optimization with the help of prior knowledge of the geometry of general thermal coagulations. The performance of the algorithm has been shown to be comparable to manual delineation of coagulations on elastograms by medical physicists (r = 0.99 for volumes of 36 radiofrequency-induced coagulations). Furthermore, the automatic algorithm applied to elastograms yielded results that agreed with manual delineation of coagulations on pathology images (r = 0.96 for the same 36 lesions). This algorithm has also been successfully applied on in vivo elastograms.

Solid breast masses: classification with computer-aided analysis of continuous US images obtained with probe compression

PURPOSE

To prospectively evaluate the accuracy of continuous ultrasonographic (US) images obtained during probe compression and computer-aided analysis for classification of biopsy-proved (reference standard) benign and malignant breast tumors.

MATERIALS AND METHODS

This study was approved by the local ethics committee, and informed consent was obtained from all included patients. Serial US images of 100 solid breast masses (60 benign and 40 malignant tumors) were obtained with US probe compression in 86 patients (mean age, 45 years; range, 20-67 years). After segmentation of tumor contours with the level-set method, three features of strain on tissue from probe compression--contour difference, shift distance, area difference--and one feature of shape--solidity-were computed. A maximum margin classifier was used to classify the tumors by using these four features. The Student t test and receiver operating characteristic curve analysis were used for statistical analysis.

RESULTS

The mean values of contour difference, shift distance, area difference, and solidity were 3.52% +/- 2.12 (standard deviation), 2.62 +/- 1.31, 1.08% +/- 0.85, and 1.70 +/- 1.85 in malignant tumors and 9.72% +/- 4.54, 5.04 +/- 2.79, 3.17% +/- 2.86, and 0.53 +/- 0.63 in benign tumors, respectively. Differences with P < .001 were statistically significant for all four features. Area under the receiver operating characteristic curve (A(Z)) values for contour difference, shift distance, area difference, and solidity were 0.88, 0.85, 0.86, and 0.79, respectively. The A(Z) value of three features of strain was significantly higher than that of the feature of shape (P < .01). The accuracy, sensitivity, specificity, and positive and negative predictive values of US classifications that were based on values for these four features were 87.0% (87 of 100), 85% (34 of 40), 88% (53 of 60), 83% (34 of 41), and 90% (53 of 59), respectively, with an A(Z) value of 0.91.

CONCLUSION

Continuous US images obtained with probe compression and computer-aided analysis can aid in classification of benign and malignant breast tumors.

Automatic ultrasound segmentation and morphology based diagnosis of solid breast tumors

Ultrasound (US) is a useful diagnostic tool to distinguish benign from malignant masses of the breast. It is a very convenient and safe diagnostic method. However, there is a considerable overlap benignancy and malignancy in ultrasonic images and interpretation is subjective. A high performance breast tumors computer-aided diagnosis (CAD) system can provide an accurate and reliable diagnostic second opinion for physicians to distinguish benign breast lesions from malignant ones. The potential of sonographic texture analysis to improve breast tumor classifications has been demonstrated. However, the texture analysis is system-dependent. The disadvantages of these systems which use texture analysis to classify tumors are they usually perform well only in one specific ultrasound system. While Morphological based US diagnosis of breast tumor will take the advantage of nearly independent to either the setting of US system and different US machines. In this study, the tumors are segmented using the newly developed level set method at first and then six morphologic features are used to distinguish the benign and malignant cases. The support vector machine (SVM) is used to classify the tumors. There are 210 ultrasonic images of pathologically proven benign breast tumors from 120 patients and carcinomas from 90 patients in the ultrasonic image database. The database contains only one image from each patient. The ultrasonic images are captured at the largest diameter of the tumor. The images are collected consecutively from August 1, 1999 to May 31, 2000; the patients' ages ranged from 18 to 64 years. Sonography is performed using an ATL HDI 3000 system with a L10-5 small part transducer. In the experiment, the accuracy of SVM with shape information for classifying malignancies is 90.95% (191/210), the sensitivity is 88.89% (80/90), the specificity is 92.5% (111/120), the positive predictive value is 89.89% (80/89), and the negative predictive value is 91.74% (111/121).

Watershed segmentation for breast tumor in 2-D sonography

Automatic contouring for breast tumors using medical ultrasound (US) imaging may assist physicians without relevant experience, in making correct diagnoses. This study integrates the advantages of neural network (NN) classification and morphological watershed segmentation to extract precise contours of breast tumors from US images. Textural analysis is employed to yield inputs to the NN to classify ultrasonic images. Autocovariance coefficients specify texture features to classify breasts imaged by US using a self-organizing map (SOM). After the texture features in sonography have been classified, an adaptive preprocessing procedure is selected by SOM output. Finally, watershed transformation automatically determines the contours of the tumor. In this study, the proposed method was trained and tested using images from 60 patients. The results of computer simulations reveal that the proposed method always identified similar contours and regions-of-interest (ROIs) to those obtained by manual contouring (by an experienced physician) of the breast tumor in ultrasonic images. As US imaging becomes more widespread, a functional automatic contouring method is essential and its clinical application is becoming urgent. Such a method provides robust and fast automatic contouring of US images. This study is not to emphasize that the automatic contouring technique is superior to the one undertaken manually. Both automatic and manual contours did not, after all, necessarily result in the same factual pathologic border. In computer-aided diagnosis (CAD) applications, automatic segmentation can save much of the time required to sketch a precise contour, with very high stability.

Three-dimensional ultrasound in margin evaluation for breast tumor excision using Mammotome

Sonographic evidence of tumor removal by Mammotome excision does not confirm histological clearance. The operator finds it hard to determine if a malignant tumor has been fully removed, leaving a safe margin in the direction of each border; that is, the spatial orientation during tumor retrieval is not well-established by naked eye under sonographic guidance. We propose a computational imaging process to extract reasonable tumor contour in pre- and postoperative data sets for sonographic guidance so that Mammotome excision can help the operator to evaluate the surgical outcome. There were five tumors in the study, including three benign and two malignant. The lesion of interest was delineated after 2-D examination was completed, then it was analyzed with 3-D breast ultrasound (US). To give a reference point for correlations between pre- and postoperative images, we used a marker tape pasted on the skin within the transducer scanning area and then the preoperative 3-D US images were obtained. Subsequently, 2-D breast US was applied during Mammotome operation. After the Mammotome procedures were finished, the postoperative 3-D US images were obtained; thus, we gained two different data sets of 3-D US images that were used for later analysis for evaluating the extension of postoperative margin status. From the results, the safe margin was not satisfactory in all directions, because the minimum differences measured by the proposed algorithm were not large enough in all five cases, and this was proved from two malignant mastectomy specimens. The experimental results representing this inadequate Mammotome excision can be visualized through the computer aid. The comparison of tumor contour and excision margin may possibly be used for small malignant tumors in the future to improve the breast-conserving surgery.