Active contour external force using vector field convolution for image segmentation

Segmentation-based BI-RADS ensemble classification of breast tumours in ultrasound images

BACKGROUND

The development of computer-aided diagnosis systems in breast cancer imaging is exponential. Since 2016, 81 papers have described the automated segmentation of breast lesions in ultrasound images using artificial intelligence. However, only two papers have dealt with complex BI-RADS classifications.

PURPOSE

This study addresses the automatic classification of breast lesions into binary classes (benign vs. malignant) and multiple BI-RADS classes based on a single ultrasonographic image. Achieving this task should reduce the subjectivity of an individual operator's assessment.

MATERIALS AND METHODS

Automatic image segmentation methods (PraNet, CaraNet and FCBFormer) adapted to the specific segmentation task were investigated using the U-Net model as a reference. A new classification method was developed using an ensemble of selected segmentation approaches. All experiments were performed on publicly available BUS B, OASBUD, BUSI and private datasets.

RESULTS

FCBFormer achieved the best outcomes for the segmentation task with intersection over union metric values of 0.81, 0.80 and 0.73 and Dice values of 0.89, 0.87 and 0.82, respectively, for the BUS B, BUSI and OASBUD datasets. Through a series of experiments, we determined that adding an extra 30-pixel margin to the segmentation mask counteracts the potential errors introduced by the segmentation algorithm. An assembly of the full image classifier, bounding box classifier and masked image classifier was the most accurate for binary classification and had the best accuracy (ACC; 0.908), F1 (0.846) and area under the receiver operating characteristics curve (AUROC; 0.871) in the BUS B and ACC (0.982), F1 (0.984) and AUROC (0.998) in the UCC BUS datasets, outperforming each classifier used separately. It was also the most effective for BI-RADS classification, with ACC of 0.953, F1 of 0.920 and AUROC of 0.986 in UCC BUS. Hard voting was the most effective method for dichotomous classification. For the multi-class BI-RADS classification, the soft voting approach was employed.

CONCLUSIONS

The proposed new classification approach with an ensemble of segmentation and classification approaches proved more accurate than most published results for binary and multi-class BI-RADS classifications.

Multi-agent medical image segmentation: A survey

During the last decades, the healthcare area has increasingly relied on medical imaging for the diagnosis of a growing number of pathologies. The different types of medical images are mostly manually processed by human radiologists for diseases detection and monitoring. However, such a procedure is time-consuming and relies on expert judgment. The latter can be influenced by a variety of factors. One of the most complicated image processing tasks is image segmentation. Medical image segmentation consists of dividing the input image into a set of regions of interest, corresponding to body tissues and organs. Recently, artificial intelligence (AI) techniques brought researchers attention with their promising results for the image segmentation automation. Among AI-based techniques are those that use the Multi-Agent System (MAS) paradigm. This paper presents a comparative study of the multi-agent approaches dedicated to the segmentation of medical images, recently published in the literature.

Image Segmentation Using Active Contours with Hessian-Based Gradient Vector Flow External Force

The gradient vector flow (GVF) model has been widely used in the field of computer image segmentation. In order to achieve better results in image processing, there are many research papers based on the GVF model. However, few models include image structure. In this paper, the smoothness constraint formula of the GVF model is re-expressed in matrix form, and the image knot represented by the Hessian matrix is included in the GVF model. Through the processing of this process, the relevant diffusion partial differential equation has anisotropy. The GVF model based on the Hessian matrix (HBGVF) has many advantages over other relevant GVF methods, such as accurate convergence to various concave surfaces, excellent weak edge retention ability, and so on. The following will prove the advantages of our proposed model through theoretical analysis and various comparative experiments.

Automatic breast tissue segmentation in MRIs with morphology snake and deep denoiser training via extended Stein's unbiased risk estimator

Accurate segmentation of the breast tissue is a significant challenge in the analysis of breast MR images, especially analysis of breast images with low contrast. Most of the existing methods for breast segmentation are semi-automatic and limited in their ability to achieve accurate results. This is because of difficulties in removing landmarks from noisy magnetic resonance images (MRI). Especially, when tumour is imaged for scanning, how to isolate the tumour region from chest will directly affect the accuracy for tumour to be detected. Due to low intensity levels and the close connection between breast and chest portion in MRIs, this study proposes an innovative, fully automatic and fast segmentation approach which combines histogram with inverse Gaussian gradient for morphology snakes, along with extended Stein's unbiased risk estimator (eSURE) applied for unsupervised learning of deep neural network Gaussian denoisers, aimed at accurate identification of landmarks such as chest and breast.

Initial Geometrical Templates with Parameter Sets for Active Contour on Skin Cancer Boundary Segmentation

For active-contour-based surgery systems, the success of skin cancer boundary segmentation depends on the initialization point of the snake model, which is a task originally performed by skillful experts, and on the parameters set for the algorithms of active contour. This paper proposes initial geometrical templates and parameter sets for the active contour on skin cancer boundary segmentation. To establish initial geometrical templates and parameter sets for the active contour, first, template candidates, which are geometrically designed by users in advance, are simply calculated based on similarity with a skin cancer boundary, and the candidate with the least difference is selected as an initial template. Initially, all candidate templates are performed before the test with some selected skin cancer samples by randomly varying needed parameters to determine parameter sets for each template. The parameter set is therefore implicitly selected as the suitable set with the selected initial template. Experiments with 227 skin cancer samples were performed based on our proposed initial templates and parameter sets, and the results show 99.46% accuracy, 97.43% sensitivity, and 99.87% specificity approximately in which accuracy, sensitivity, and specificity were improved by 0.26%, 0.36%, and 0.26%, respectively, compared with the conventional method.

Image segmentation for neuroscience: lymphatics

A recent discovery in neuroscience prompts the need for innovation in image analysis. Neuroscientists have discovered the existence of meningeal lymphatic vessels in the brain and have shown their importance in preventing cognitive decline in mouse models of Alzheimer’s disease. With age, lymphatic vessels narrow and poorly drain cerebrospinal fluid, leading to plaque accumulation, a marker for Alzheimer’s disease. The detection of vessel boundaries and width are performed by hand in current practice and thereby suffer from high error rates and potential observer bias. The existing vessel segmentation methods are dependent on user-defined initialization, which is time-consuming and difficult to achieve in practice due to high amounts of background clutter and noise. This work proposes a level set segmentation method featuring hierarchical matting, LyMPhi, to predetermine foreground and background regions. The level set force field is modulated by the foreground information computed by matting, while also constraining the segmentation contour to be smooth. Segmentation output from this method has a higher overall Dice coefficient and boundary F1-score compared to that of competing algorithms. The algorithms are tested on real and synthetic data generated by our novel shape deformation based approach. LyMPhi is also shown to be more stable under different initial conditions as compared to existing level set segmentation methods. Finally, statistical analysis on manual segmentation is performed to prove the variation and disagreement between three annotators.

Geodesic Paths for Image Segmentation With Implicit Region-Based Homogeneity Enhancement

Minimal paths are regarded as a powerful and efficient tool for boundary detection and image segmentation due to its global optimality and the well-established numerical solutions such as fast marching method. In this paper, we introduce a flexible interactive image segmentation model based on the Eikonal partial differential equation (PDE) framework in conjunction with region-based homogeneity enhancement. A key ingredient in the introduced model is the construction of local geodesic metrics, which are capable of integrating anisotropic and asymmetric edge features, implicit region-based homogeneity features and/or curvature regularization. The incorporation of the region-based homogeneity features into the metrics considered relies on an implicit representation of these features, which is one of the contributions of this work. Moreover, we also introduce a way to build simple closed contours as the concatenation of two disjoint open curves. Experimental results prove that the proposed model indeed outperforms state-of-the-art minimal paths-based image segmentation approaches.

Magnetostatic Active Contour Model with Classification Method of Sparse Representation

The active contour model is widely used to segment images. For the classical magnetostatic active contour (MAC) model, the magnetic field is computed based on the detected points by using an edge detector. However, noise and nontarget points are always detected. Thus, MAC is nonrobust to noise and the extracted objects may be deviant from the real objects. In this paper, a magnetostatic active contour model with a classification method of sparse representation is proposed. First, rough edge information is obtained with some edge detectors. Second, the extracted edge contours are divided into two parts by sparse classification, that is, the target object part and the redundant part. Based on the classified target points, a new magnetic field is generated, and contours evolve with MAC to extract the target objects. Experimental results show that the proposed model could decrease the influence of noise and robust segmentation results could be obtained.

Analysis of in vivo uterine peristalsis in the non-pregnant female mouse

Uterine peristalsis due to spontaneous contractions of the myometrial smooth muscles has important roles in pre-implantation processes of intra-uterine sperm transport to the fertilization site, and then embryo transport to the implantation sites. We developed a new objective methodology to study in vivo uterine peristalsis in female mice during the pro-oestrus phase. The acquisition procedure of the uterine organ is remote without interfering with the organ function. The uniqueness of the new approach is that video images of physiological pattern were converted using image processing and new algorithms to biological time-dependent signals that can be processed with existing algorithms for signal processing. Using this methodology we found that uterine peristalsis in the pro-oestrus mouse is in the range of 0.008-0.029 Hz, which is about one contraction per minute and with fairly symmetric contractions that occasionally propagate caudally. This rate of contractions is similar to that of human uterine peristalsis acquired in vivo, which is important information for a popular animal model.

Sparse Multi-Bending Snakes

Active contour models are one of the most emblematic algorithms of computer vision. Their strong theoretical foundations and high user interoperability turned them into a reference approach for object segmentation and tracking tasks. A high number of modifications have already been proposed in order to overcome the known problems of traditional snakes, such as initialization dependence and poor convergence to concavities. In this paper, we address the scenario where the user wants to segment an object that has multiple dynamic regions but some of them do not correspond to the true object boundary. We propose a novel parametric active contour model, the Sparse Multi-Bending snake, which is capable of dividing the contour into a set of contiguous regions with different bending properties. We derive a new energy function that induces such behavior and presents a group optimization strategy that can be used to find the optimal bending resistance parameter for each point of the contour. We show the flexibility of our model in a set of synthetic images. In addition, we consider two real applications, lung segmentation in Computerized Tomography data and hand segmentation in depth images. We show how the proposed method is able to improve the segmentations obtained in both applications, when compared with other active contour models.

Deriving external forces via convolutional neural networks for biomedical image segmentation

Active contours, or snakes, are widely applied on biomedical image segmentation. They are curves defined within an image domain that can move to object boundaries under the influence of internal forces and external forces, in which the internal forces are generally computed from curves themselves and external forces from image data. Designing external forces properly is a key point with active contour algorithms since the external forces play a leading role in the evolution of active contours. One of most popular external forces for active contour models is gradient vector flow (GVF). However, GVF is sensitive to noise and false edges, which limits its application area. To handle this problem, in this paper, we propose using GVF as reference to train a convolutional neural network to derive an external force. The derived external force is then integrated into the active contour models for curve evolution. Three clinical applications, segmentation of optic disk in fundus images, fluid in retinal optical coherence tomography images and fetal head in ultrasound images, are employed to evaluate the proposed method. The results show that the proposed method is very promising since it achieves competitive performance for different tasks compared to the state-of-the-art algorithms.

Initialization of active contours for segmentation of breast cancer via fusion of ultrasound, Doppler, and elasticity images

Active contours (snakes) are an efficient method for segmentation of ultrasound (US) images of breast cancer. However, the method produces inaccurate results if the seeds are initialized improperly (far from the true boundaries and close to the false boundaries). Therefore, we propose a novel initialization method based on the fusion of a conventional US image with elasticity and Doppler images. The proposed fusion method (FM) has been tested against four state-of-the-art initialization methods on 90 ultrasound images from a database collected by the Thammasat University Hospital of Thailand. The ground truth was hand-drawn by three leading radiologists of the hospital. The reference methods are: center of divergence (CoD), force field segmentation (FFS), Poisson Inverse Gradient Vector Flow (PIG), and quasi-automated initialization (QAI). A variety of numerical tests proves the advantages of the FM. For the raw US images, the percentage of correctly initialized contours is: FM-94.2%, CoD-0%, FFS-0%, PIG-26.7%, QAI-42.2%. The percentage of correctly segmented tumors is FM-84.4%, CoD-0%, FFS-0%, PIG-16.67%, QAI-22.44%. For reduced field of view US images, the percentage of correctly initialized contours is: FM-94.2%, CoD-0%, FFS-0%, PIG-65.6%, QAI-67.8%. The correctly segmented tumors are FM-88.9%, CoD-0%, FFS-0%, PIG-48.9%, QAI-44.5%. The accuracy, in terms of the average Hausdorff distance, is respectively 2.29 pixels, 33.81, 34.71, 7.7, and 8.4, whereas in terms of the Jaccard index, it is 0.9, 0.18, 0.19, 0.63, and 0.48.

Layer boundary evolution method for macular OCT layer segmentation

Optical coherence tomography (OCT) is used to produce high resolution depth images of the retina and is now the standard of care for in-vivo ophthalmological assessment. It is also increasingly being used for evaluation of neurological disorders such as multiple sclerosis (MS). Automatic segmentation methods identify the retinal layers of the macular cube providing consistent results without intra- and inter-rater variation and is faster than manual segmentation. In this paper, we propose a fast multi-layer macular OCT segmentation method based on a fast level set method. Our framework uses contours in an optimized approach specifically for OCT layer segmentation over the whole macular cube. Our algorithm takes boundary probability maps from a trained random forest and iteratively refines the prediction to subvoxel precision. Evaluation on both healthy and multiple sclerosis subjects shows that our method is statistically better than a state-of-the-art graph-based method.

Vector Field Convolution-Based B-Spline Deformation Model for 3D Segmentation of Cartilage in MRI

In this paper, a novel 3D vector field convolution (VFC)-based B-spline deformation model is proposed for accurate and robust cartilage segmentation. Firstly, the anisotropic diffusion method is utilized for noise reduction, and the Sinc interpolation method is employed for resampling. Then, to extract the rough cartilage, features derived from

Computational Analysis of Cell Dynamics in Videos with Hierarchical-Pooled Deep-Convolutional Features

Computational analysis of cellular appearance and its dynamics is used to investigate physiological properties of cells in biomedical research. In consideration of the great success of deep learning in video analysis, we first introduce two-stream convolutional networks (ConvNets) to automatically learn the biologically meaningful dynamics from raw live-cell videos. However, the two-stream ConvNets lack the ability to capture long-range video evolution. Therefore, a novel hierarchical pooling strategy is proposed to model the cell dynamics in a whole video, which is composed of trajectory pooling for short-term dynamics and rank pooling for long-range ones. Experimental results demonstrate that the proposed pipeline effectively captures the spatiotemporal dynamics from the raw live-cell videos and outperforms existing methods on our cell video database.

SDL: Saliency-Based Dictionary Learning Framework for Image Similarity

In image classification, obtaining adequate data to learn a robust classifier has often proven to be difficult in several scenarios. Classification of histological tissue images for health care analysis is a notable application in this context due to the necessity of surgery, biopsy or autopsy. To adequately exploit limited training data in classification, we propose a saliency guided dictionary learning method and subsequently an image similarity technique for histo-pathological image classification. Salient object detection from images aids in the identification of discriminative image features. We leverage the saliency values for the local image regions to learn a dictionary and respective sparse codes for an image, such that the more salient features are reconstructed with smaller error. The dictionary learned from an image gives a compact representation of the image itself and is capable of representing images with similar content, with comparable sparse codes. We employ this idea to design a similarity measure between a pair of images, where local image features of one image, are encoded with the dictionary learned from the other and vice versa. To effectively utilize the learned dictionary, we take into account the contribution of each dictionary atom in the sparse codes to generate a global image representation for image comparison. The efficacy of the proposed method was evaluated using three tissue data sets that consist of mammalian kidney, lung and spleen tissue, breast cancer, and colon cancer tissue images. From the experiments, we observe that our methods outperform the state of the art with an increase of 14.2% in the average classification accuracy over all data sets.

Robust Evolution Method of Active Contour Models and Application in Segmentation of Image Sequence

Active contour models are widely used in image segmentation. In order to obtain ideal object boundary, researchers utilize various information to define new models for image segmentation. However, the models could not meet all scenes of image. In this paper, we propose a block evolution method to improve the robustness of contour evolution. A block matrix is consisted of contours of former iterations and contours of shape prior, and a nuclear norm of the matrix is a measure of the similarity of these shapes. The constraint of the nuclear norm minimization is imposed on the evolution of active contour models, which could avoid large deformation of the adjacent curves and keep the shape conformability of contour in the evolution. The shape prior can be integrated into the block evolution method, which is effective in dealing with missing features of images and noise. The proposed method can be applied to image sequence segmentation. Experiments demonstrate that the proposed method improves the robust performance of active contour models and can increase the flexibility of applications in image sequence segmentation.

The first MICCAI challenge on PET tumor segmentation

INTRODUCTION

Automatic functional volume segmentation in PET images is a challenge that has been addressed using a large array of methods. A major limitation for the field has been the lack of a benchmark dataset that would allow direct comparison of the results in the various publications. In the present work, we describe a comparison of recent methods on a large dataset following recommendations by the American Association of Physicists in Medicine (AAPM) task group (TG) 211, which was carried out within a MICCAI (Medical Image Computing and Computer Assisted Intervention) challenge.

MATERIALS AND METHODS

Organization and funding was provided by France Life Imaging (FLI). A dataset of 176 images combining simulated, phantom and clinical images was assembled. A website allowed the participants to register and download training data (n = 19). Challengers then submitted encapsulated pipelines on an online platform that autonomously ran the algorithms on the testing data (n = 157) and evaluated the results. The methods were ranked according to the arithmetic mean of sensitivity and positive predictive value.

RESULTS

Sixteen teams registered but only four provided manuscripts and pipeline(s) for a total of 10 methods. In addition, results using two thresholds and the Fuzzy Locally Adaptive Bayesian (FLAB) were generated. All competing methods except one performed with median accuracy above 0.8. The method with the highest score was the convolutional neural network-based segmentation, which significantly outperformed 9 out of 12 of the other methods, but not the improved K-Means, Gaussian Model Mixture and Fuzzy C-Means methods.

CONCLUSION

The most rigorous comparative study of PET segmentation algorithms to date was carried out using a dataset that is the largest used in such studies so far. The hierarchy amongst the methods in terms of accuracy did not depend strongly on the subset of datasets or the metrics (or combination of metrics). All the methods submitted by the challengers except one demonstrated good performance with median accuracy scores above 0.8.

Segmentation of the right ventricle in four chamber cine cardiac MR images using polar dynamic programming

The four chamber plane is currently underutilized in the right ventricular segmentation community. Four chamber information can be useful to determine ventricular short axis stacks and provide a rough estimate of the right ventricle in short axis stacks. In this study, we develop and test a semi-automated technique for segmenting the right ventricle in four chamber cine cardiac magnetic resonance images. The three techniques that use minimum cost path algorithms were used. The algorithms are: Dijkstra's shortest path algorithm (Dijkstra), an A* algorithm that uses length, curvature and torsion into an active contour model (ALCT), and a variation of polar dynamic programming (PDP). The techniques are evaluated against the expert traces using 175 cardiac images from 7 patients. The evaluation first looks at mutual overlap metrics and then focuses on clinical measures such as fractional area change (FAC). The mean mutual overlap between the physician's traces ranged from 0.85 to 0.88. Using as reference physician 1's landmarks and traces (i.e., comparing the traces from physician 1 to the semi-automated segmentation using physician 1's landmarks), the PDP algorithm has a mean mutual overlap of 0.8970 compared to 0.8912 for ALCT and 0.8879 for Dijkstra. The mean mutual overlap between the BP regions generated by physician 1 and physician 2 landmarks are 0.9674, 0.9605 and 0.9531 for PDP, ALCT and Dijkstra, respectively. The FAC correlation coefficient between the physician's traces ranged from 0.73 to 0.93.

Automatic initialization for active contour model in breast cancer detection utilizing conventional ultrasound and Color Doppler

Regular examination of breasts may prevent and help to cure because breast cancer is treatable when it is detected early. Therefore, a breast cancer screening modality being sensitivity and cost-effective like ultrasonic imaging modality (US), is strongly required. In addition, the combination of a conventional US and its adjunct, Color Doppler has been proved for decreasing the rate of false-positive in breast cancer diagnosis. Thus, combination of these imaging modalities in a breast cancer segmentation would provide some benefits as well. An effective method for feature segmentation, active contour model has been widely utilized for decades. A crucial stage that affects the performance of active contour model is the initialization. This paper proposes a novel method for an automatic initialization of active contour model designed specifically for US-based imaging modalities. The method estimates an initial contour by utilizing the fusion of conventional US and Color Doppler. Examples and comparisons with three state-of-the-art automatic initialization methods are demonstrated to present the advantages of the proposed method. The evaluation results show high accuracy of initialization as well as fast convergence to features of interest.

MyoVision: software for automated high-content analysis of skeletal muscle immunohistochemistry

Analysis of skeletal muscle cross sections is an important experimental technique in muscle biology. Many aspects of immunohistochemistry and fluorescence microscopy can now be automated, but most image quantification techniques still require extensive human input, slowing progress and introducing the possibility of user bias. MyoVision is a new software package that was developed to overcome these limitations. The software improves upon previously reported automatic techniques and analyzes images without requiring significant human input and correction. When compared with data derived by manual quantification, MyoVision achieves an accuracy of ≥94% for basic measurements such as fiber number, fiber type distribution, fiber cross-sectional area, and myonuclear number. Scientists can download the software free from www.MyoVision.org and use it to automate the analysis of their own experimental data. This will improve the efficiency and consistency of the analysis of muscle cross sections and help to reduce the burden of routine image quantification in muscle biology. NEW & NOTEWORTHY Scientists currently analyze images of immunofluorescently labeled skeletal muscle using time-consuming techniques that require sustained human supervision. As well as being inefficient, these techniques can increase variability in studies that quantify morphological adaptations of skeletal muscle at the cellular level. MyoVision is new software that overcomes these limitations by performing high-content analysis of muscle cross sections with minimal manual input. It is open source and freely available.

Automatic measurement of the sinus of Valsalva by image analysis

BACKGROUND AND OBJECTIVES

Despite the importance of the morphology of the sinus of Valsalva in the behavior of heart valves and the proper irrigation of coronary arteries, the study of these sinuses from medical imaging is still limited to manual radii measurements. This paper aims to present an automatic method to measure the sinuses of Valsalva on medical images, more specifically on cine MRI and Xray CT.

METHODS

This paper introduces an enhanced method to automatically localize and extract each sinus of Valsalva edge and its relevant points. Compared to classical active contours, this new image approach enhances the edge extraction of the Sinus of Valsalva. Our process not only allows image segmentation but also a complex study of the considered region including morphological classification, metrological characterization, valve tracking and 2D modeling.

RESULTS

The method was successfully used on single or multiplane cine MRI and aortic CT angiographies. The localization is robust and the proposed edge extractor is more efficient than the state-of-the-art methods (average success rate for MRI examinations=84% ± 24%, average success rate for CT examinations=89% ± 11%). Moreover, deduced measurements are close to manual ones.

CONCLUSIONS

The software produces accurate measurements of the sinuses of Valsalva. The robustness and the reproducibility of results will help for a better understanding of sinus of Valsalva pathologies and constitutes a first step to the design of complex prostheses adapted to each patient.

Error performances of a model-based biplane fluoroscopic system for tracking knee prosthesis during treadmill gait task

Roentgen stereophotogrammetry analysis technique allows an accurate measurement of knee joint prosthesis position and orientation using two X-ray images. Although this technique is used generally during static procedure, it is possible to use it with a biplane fluoroscopic system to measure the prosthesis kinematics during functional tasks (e.g., gait, squat, jump) performed in a laboratory environment. However, the performance of the system in terms of errors for the measurements and the model-based matching algorithm are not well known for dynamic tasks such as walking. The goal of this study was to estimate the static and dynamic errors of a model-based biplane fluoroscopic system for a treadmill gait task and analyze the error performance according to the speed and location of the knee joint prosthesis relative to X-ray sources. The results show a static maximum error (RMSE) of 0.13° for orientation and 0.06 mm for position for prosthesis components. The dynamic errors were different for each axis of the acquisition system and each prosthesis component. The largest dynamic error was along the vertical axis for the position (RMSE = 2.42 mm) and along the medio-lateral axis (perpendicular to movement) for the orientation (RMSE = 0.95°). As expected, the error depends on the distance between the prosthesis and the source in the acquisition system as well as the linear and angular velocity of the movement. The most accurate dynamic measure was around the centroid of the acquisition system, while kinematics measurements close to the X-rays detectors gave the worst errors.

Segmentation of medical images using mean value guided contour

Partial differential equation-based (PDE-based) methods are extensively used in image segmentation, especially in contour models. Difficulties associated with the boundaries, namely troubles with developing initialization, inadequate convergence to boundary concavities, and difficulties connected to saddle points and stationary points of active contours make the contour models suffer from a feeble performance of referring to complex geometries. The present paper is designed to take advantage of mean value theorem rather than minimizing energy function for contours. It is efficiently capable of resolving above-mentioned problems by applying this theorem to the edge map gradient vectors, which is calculated from the image. Since the contour is computed in a straightforward manner from an edge map instead of force balance equation, it varies from other contour-based image segmentation methods. To illustrate the ability of the proposed model in complex geometries and ruptures, several experiments were also provided to validate the model. The experiments' results demonstrated that the proposed method, which is called mean value guided contour (MVGC), is capable of repositioning contours into boundary concavities and has suitable forcefulness in complex geometries.

Automatic construction of statistical shape models using deformable simplex meshes with vector field convolution energy

BACKGROUND

In the active shape model framework, principal component analysis (PCA) based statistical shape models (SSMs) are widely employed to incorporate high-level a priori shape knowledge of the structure to be segmented to achieve robustness. A crucial component of building SSMs is to establish shape correspondence between all training shapes, which is a very challenging task, especially in three dimensions.

METHODS

We propose a novel mesh-to-volume registration based shape correspondence establishment method to improve the accuracy and reduce the computational cost. Specifically, we present a greedy algorithm based deformable simplex mesh that uses vector field convolution as the external energy. Furthermore, we develop an automatic shape initialization method by using a Gaussian mixture model based registration algorithm, to derive an initial shape that has high overlap with the object of interest, such that the deformable models can then evolve more locally. We apply the proposed deformable surface model to the application of femur statistical shape model construction to illustrate its accuracy and efficiency.

RESULTS

Extensive experiments on ten femur CT scans show that the quality of the constructed femur shape models via the proposed method is much better than that of the classical spherical harmonics (SPHARM) method. Moreover, the proposed method achieves much higher computational efficiency than the SPHARM method.

CONCLUSIONS

The experimental results suggest that our method can be employed for effective statistical shape model construction.

Objective Analysis of Vaginal Ultrasound Video Clips for Exploring Uterine Peristalsis Post Vaginal and Cesarean Section Deliveries

The nonpregnant uterus is characterized by cyclic contractions that assist in sperm transport to the fallopian tube, embryo transport to implantation site, and expulsion of menstrual debris. The effect of post-Cesarean section (CS) scar on uterine peristalsis is unclear, while worldwide the prevalence of CS deliveries is increasing. In this study, we developed a new objective method for analysis of dynamic characteristics of the nonpregnant uterus from transvaginal ultrasound (TVUS) recordings when the uterine cavity is not clearly observed, as may be the case in post-CS uteri. The method of active contours was utilized to detect the contours of the endometrium-myometrium interface (EMI) from sagittal cross-section TVUS images of nonpregnant uteri. The contours were straightened along the uterus centerline and registered with respect to the fundal end in order to reduce the noise due to movements of the physician and the participant. A dynamic analysis was conducted on these time-dependent contours in order to explore the frequency and amplitude of the EMI motility. The analysis was conducted on TVUS video clips from 12 nonpregnant participants, 7 post-CS and 5 controls. The frequencies of the EMI motility was 0.010 to 0.064 Hz at days 8 to 17 in the control participants and 0.014 to 0.073 Hz at days 9 to 15 in post-CS participants. The maximal amplitude of motility was 0.67 to 2.00 mm and 0.48 to 2.58 mm for the control and post-CS participants, respectively. In this preliminary study, we have not observed significant difference between the EMI motility of healthy and post-CS uteri.

Multiresolution Subdivision Snakes

We present a new family of snakes that satisfy the property of multiresolution by exploiting subdivision schemes. We show in a generic way how to construct such snakes based on an admissible subdivision mask. We derive the necessary energy formulations and provide the formulas for their efficient computation. Depending on the choice of the mask, such models have the ability to reproduce trigonometric or polynomial curves. They can also be designed to be interpolating, a property that is useful in user-interactive applications. We provide explicit examples of subdivision snakes and illustrate their use for the segmentation of bioimages. We show that they are robust in the presence of noise and provide a multiresolution algorithm to enlarge their basin of attraction, which decreases their dependence on initialization compared to singleresolution snakes. We show the advantages of the proposed model in terms of computation and segmentation of structures with different sizes.

Semi-automated myocardial segmentation of bright blood multi-gradient echo images improves reproducibility of myocardial contours and T2* determination

Objectives

Early detection of iron loading is affected by the reproducibility of myocardial contour assessment. A novel semi-automatic myocardial segmentation method is presented on contrast-optimized composite images and compared to the results of manual drawing.

Materials and methods

Fifty-one short-axis slices at basal, mid-ventricular and apical locations from 17 patients were acquired by bright blood multi-gradient echo MRI. Four observers produced semi-automatic and manual myocardial contours on contrast-optimized composite images. The semi-automatic segmentation method relies on vector field convolution active contours to generate the endocardial contour. After creating radial pixel clusters on the myocardial wall, a combination of pixel-wise coefficient of variance (CoV) assessment and k-means clustering establishes the epicardial contour for each segment.

Results

Compared to manual drawing, semi-automatic myocardial segmentation lowers the variability of T2* quantification within and between observers (CoV of 12.05 vs. 13.86% and 14.43 vs. 16.01%) by improving contour reproducibility (P < 0.001). In the presence of iron loading, semi-automatic segmentation also lowers the T2* variability within and between observers (CoV of 13.14 vs. 15.19% and 15.91 vs. 17.28%).

Conclusion

Application of semi-automatic myocardial segmentation on contrast-optimized composite images improves the reproducibility of T2* quantification.

Dynamic Programming Using Polar Variance for Image Segmentation

When using polar dynamic programming (PDP) for image segmentation, the object size is one of the main features used. This is because if size is left unconstrained the final segmentation may include high-gradient regions that are not associated with the object. In this paper, we propose a new feature, polar variance, which allows the algorithm to segment the objects of different sizes without the need for training data. The polar variance is the variance in a polar region between a user-selected origin and a pixel we want to analyze. We also incorporate a new technique that allows PDP to segment complex shapes by finding low-gradient regions and growing them. The experimental analysis consisted on comparing our technique with different active contour segmentation techniques on a series of tests. The tests consisted on robustness to additive Gaussian noise, segmentation accuracy with different grayscale images and finally robustness to algorithm-specific parameters. Experimental results show that our technique performs favorably when compared with other segmentation techniques.

Local Transverse-Slice-Based Level-Set Method for Segmentation of 3-D High-Frequency Ultrasonic Backscatter From Dissected Human Lymph Nodes

OBJECTIVE

To detect metastases in freshly excised human lymph nodes (LNs) using three-dimensional (3-D), high-frequency, quantitative ultrasound (QUS) methods, the LN parenchyma (LNP) must be segmented to preclude QUS analysis of data in regions outside the LNP and to compensate ultrasound attenuation effects due to overlying layers of LNP and residual perinodal fat (PNF).

METHODS

After restoring the saturated radio-frequency signals from PNF using an approach based on smoothing cubic splines, the three regions, i.e., LNP, PNF, and normal saline (NS), in the LN envelope data are segmented using a new, automatic, 3-D, three-phase, statistical transverseslice-based level-set (STS-LS) method that amends Lankton's method. Due to ultrasound attenuation and focusing effects, the speckle statistics of the envelope data vary with imaged depth. Thus, to mitigate depth-related inhomogeneity effects, the STS-LS method employs gamma probabilitydensity functions to locally model the speckle statistics within consecutive transverse slices.

RESULTS

Accurate results were obtained on simulated data. On a representative dataset of 54 LNs acquired from colorectal-cancer patients, the Dice similarity coefficient for LNP, PNF, and NS were 0.938 ± 0.025, 0.832 ± 0.086, and 0.968 ± 0.008, respectively, when compared to expert manual segmentation.

CONCLUSION

The STS-LS outperforms the established methods based on global and local statistics in our datasets and is capable of accurately handling the depth-dependent effects due to attenuation and focusing.

SIGNIFICANCE

This advance permits the automatic QUS-based cancer detection in the LNs. Furthermore, the STS-LS method could potentially be used in a wide range of ultrasound-imaging applications suffering from depth-dependent effects.

An Efficient Approach for Automated Mass Segmentation and Classification in Mammograms

Breast cancer is becoming a leading death of women all over the world; clinical experiments demonstrate that early detection and accurate diagnosis can increase the potential of treatment. In order to improve the breast cancer diagnosis precision, this paper presents a novel automated segmentation and classification method for mammograms. We conduct the experiment on both DDSM database and MIAS database, firstly extract the region of interests (ROIs) with chain codes and using the rough set (RS) method to enhance the ROIs, secondly segment the mass region from the location ROIs with an improved vector field convolution (VFC) snake and following extract features from the mass region and its surroundings, and then establish features database with 32 dimensions; finally, these features are used as input to several classification techniques. In our work, the random forest is used and compared with support vector machine (SVM), genetic algorithm support vector machine (GA-SVM), particle swarm optimization support vector machine (PSO-SVM), and decision tree. The effectiveness of our method is evaluated by a comprehensive and objective evaluation system; also, Matthew's correlation coefficient (MCC) indicator is used. Among the state-of-the-art classifiers, our method achieves the best performance with best accuracy of 97.73%, and the MCC value reaches 0.8668 and 0.8652 in unique DDSM database and both two databases, respectively. Experimental results prove that the proposed method outperforms the other methods; it could consider applying in CAD systems to assist the physicians for breast cancer diagnosis.

Hermite Snakes With Control of Tangents

We introduce a new model of parametric contours defined in a continuous fashion. Our curve model relies on Hermite spline interpolation and can easily generate curves with sharp discontinuities; it also grants direct access to the tangent at each location. With these two features, the Hermite snake distinguishes itself from classical spline-snake models and allows one to address certain bioimaging problems in a more efficient way. More precisely, the Hermite snake construction allows introducing sharp corners in the snake curve and designing directional energy functionals relying on local orientation information in the input image. Using the formalism of spline theory, the model is shown to meet practical requirements such as invariance to affine transformations and good approximation properties. Finally, the dependence on initial conditions and the robustness to the noise is studied on synthetic data in order to validate our Hermite snake model, and its usefulness is illustrated on real biological images acquired using brightfield, phase-contrast, differential-interference-contrast, and scanning-electron microscopy.

Efficient Shape Priors for Spline-Based Snakes

Parametric active contours are an attractive approach for image segmentation, thanks to their computational efficiency. They are driven by application-dependent energies that reflect the prior knowledge on the object to be segmented. We propose an energy involving shape priors acting in a regularization-like manner. Thereby, the shape of the snake is orthogonally projected onto the space that spans the affine transformations of a given shape prior. The formulation of the curves is continuous, which provides computational benefits when compared with landmark-based (discrete) methods. We show that this approach improves the robustness and quality of spline-based segmentation algorithms, while its computational overhead is negligible. An interactive and ready-to-use implementation of the proposed algorithm is available and was successfully tested on real data in order to segment Drosophila flies and yeast cells in microscopic images.

Divergence of Gradient Convolution: Deformable Segmentation With Arbitrary Initializations

In this paper, we propose a unified approach to deformable model-based segmentation. The fundamental force field of the proposed method is based on computing the divergence of a gradient convolution field (GCF), which makes the full use of directional information of the image gradient vectors and their interactions across image domain. However, instead of directly using such a vector field for deformable segmentation as in the conventional approaches, we derive a more salient representation for contour evolution, and very importantly, we demonstrate that this representation of image force field not only leads to global minimum through convex relaxation but also can achieve the same result using the conventional gradient descent with an intrinsic regularization. Thus, the proposed method can handle arbitrary initializations. The proposed external force field for deformable segmentation has both edge-based properties in that the GCF is computed from image gradients, and the region-based attributes since its divergence can be treated as a region indication function. Moreover, nonlinear diffusion can be conveniently applied to GCF to improve its performance in dealing with noise interference. We also show the extension of GCF from 2D to 3D. In comparison to the state-of-the-art deformable segmentation techniques, the proposed method shows greater flexibility in model initialization and optimization realization, as well as better performance toward noise interference and appearance variation.