Level Set Evolution Method Research Articles

Near-infrared vascular image segmentation using improved level set method

Near-infrared vascular images are crucial for vascular disease diagnosis and treatment. To solve the problems of difficult segmentation and inaccurate computation issues, we propose a series of image processing methods. First, the images are preprocessed, including background removal, contrast stretching, and noise suppression. After that, the images are enhanced with a two-stage image enhancement method, which successfully combines the benefits of convolutional neural network and traditional image enhancement method. Finally, the images are segmented by an Adaptive Prior Shape Level Set Evolution (APSLSE) method, which effectively improves the common problems of initial contour sensitivity and unidirectional variation of area terms. A shape restriction item is designed to the level set function to improve its suitability for the segmentation of blood vessels. A dataset of 360 images is collected for the research and validation of above algorithms. Extensive experimental results show that the proposed algorithms can effectively process the poor quality near-infrared blood vessel images and segment their vascular shape. Compared the segmentation results with those manually labelled by experts, the False Negative Rate (FNR) is 0.1930 and False Positive Rate (FPR) is 0.04633.

AW-SDRLSE: Adaptive Weighting and Scalable Distance Regularized Level Set Evolution for Lymphoma Segmentation on PET Images.

Accurate lymphoma segmentation on Positron Emission Tomography (PET) images is of great importance for medical diagnoses, such as for distinguishing benign and malignant. To this end, this paper proposes an adaptive weighting and scalable distance regularized level set evolution (AW-SDRLSE) method for delineating lymphoma boundaries on 2D PET slices. There are three important characteristics with respect to AW-SDRLSE: 1) A scalable distance regularization term is proposed and a parameter q can control the contour's convergence rate and precision in theory. 2) A novel dynamic annular mask is proposed to calculate mean intensities of local interior and exterior regions and further define the region energy term. 3) As the level set method is sensitive to parameters, we thus propose an adaptive weighting strategy for the length and area energy terms using local region intensity and boundary direction information. AW-SDRLSE is evaluated on 90 cases of real PET data with a mean Dice coefficient of 0.8796. Comparative results demonstrate the accuracy and robustness of AW-SDRLSE as well as its performance advantages as compared with related level set methods. In addition, experimental results indicate that AW-SDRLSE can be a fine segmentation method for improving the lymphoma segmentation results obtained by deep learning (DL) methods significantly.

A morphological approach to piecewise constant active contour model incorporated with the geodesic edge term

Traditional level set-based image segmentation method has to solve the level set evolution equation which is the Euler–Lagrange equation of the energy functional defined on the image domain. Solving level set evolution equation is very time-consuming, and reinitialization is usually needed. The level set evolution equation can also be solved by mathematical morphology. The morphological implementation is very simple, fast and stable. The piecewise constant active contour model incorporated with the geodesic edge term is a hybrid active contour model which combines two active contour models which are active contour model without edges and geodesic active contour model. In this paper, the mathematical morphology-based level set evolution method is applied to the piecewise constant active contour model incorporated with the geodesic edge term. The curvature morphological operator is also improved. Experimental results show that, compared with the original piecewise constant active contour model incorporated the geodesic edge term, the new mathematical morphology-based model can segment images more accurately and there are significant gains in simplicity, speed and stability. The new mathematical morphology-based model is also compared to morphological piecewise constant active contour model, morphological geodesic active contour model, traditional piecewise constant active contour model and two other active contour models. Results show that the proposed method gets the segmentation result with faster speed and higher accuracy.

IVUS images segmentation using spatial fuzzy clustering and hierarchical level set evolution

The detection of the lumen and media-adventitia (MA) borders in intravascular ultrasound (IVUS) images is crucial for quantifying plaque burdens. The challenge of the segmentation work mainly roots in various artifacts in the image. Most of the published methods involve the establishment of complex models but do not behave well on images with artifacts. In this study, aiming at automatically delineating borders in IVUS frames acquired by 20 MHz ultrasound probes, we present a fuzzy clustering-initialized hierarchical level set evolution (FC-HLSE) method. A cluster selection strategy based on the spatial fuzzy c-means (FCM) is proposed to generate the initial value and regularization term of the level set evolution (LSE). The contour convergence splits into two LSE steps between which an ingenious contour extraction (consisting of the morphological processing, the seek and linear interpolation, the gradient-based and circular fitting-based refinement) is carried out. We evaluate the proposed methodology on the publicly available 435 images by comparing auto-segmented results with the ground truth. The performance of the method is quantified using the Jaccard measure (JM), the Hausdorff distance (HD), the percentage of area difference (PAD), the linear regression and Bland-Altman analysis. Results reveal that our method can handle images with or without artifacts. The algorithm is able to extract the lumen/MA border with the JM of 0.90/0.89, the HD of 0.31/0.40 mm, the PAD of 0.07/0.08 in average, which is better in some cases compared with several state-of-the-art methods.

Extremely Optimized DRLSE Method and Its Application to Image Segmentation

This paper proposes an extremely optimized distance regularized level set evolution (EO-DRLSE) method for image segmentation applications within the framework of level set method (LSM), which combines various types of local statistical features and adopts an adaptive regularization strategy which specifically serves for zero level set. Firstly, a variable coefficient for the contour length regularization term is designed based on the local normalized entropy, which can adaptively refreshes its value with the change of local disturbance characteristics of the image. Secondly, based on the local fitting means, the constant coefficient of the region term of the original DRLSE model is modified as an adaptive variable with its value changes as the local fitting mean changes. Thirdly, an improved edge stop function is constructed based on the local fitting variances, which enables the evolution process to maintain a considerable evolution speed in the non-target noise interference position without stopping. Obviously, this gives our algorithm a particularly strong noise suppression ability. Fourthly, an additional regularization term that only deals with the zero level curve is added besides the original regularization scheme (the whole level set function (LSF) is regularized). The extended combined regularization strategy further enhances the noise suppression ability and the adaptability to weak edges of the proposed model. Fifthly, two implementation strategies named morphological snakes and significant target detection mechanism are used to ensure the rapidity and automation of the evolution process. The extensive experiments on a large variety of synthetic and real images show that the proposed algorithm achieves excellent performance in terms of accuracy of segmentation results, rapidity of evolution process, robustness against noise, adaptability to weak target edges. In addition, the factors that can have a key impact on segmentation performance are also analyzed in depth.

Semi–Automatic Corpus Callosum Segmentation and 3D Visualization Using Active Contour Methods

Accurate 3D computer models of the brain, and also of parts of its structure such as the corpus callosum (CC) are increasingly used in routine clinical diagnostics. This study presents comparative research to assess the utility and performance of three active contour methods (ACMs) for segmenting the CC from magnetic resonance (MR) images of the brain, namely: an edge-based active contour model using an inflation/deflation force with a damping coefficient (EM), the Selective Binary and Gaussian Filtering Regularized Level Set (SBGFRLS) method and the Distance Regularized Level Set Evolution (DRLSE) method. The pre-processing methods applied during research work were to improve the contrast, reduce noise and thus help segment the CC better. In this project, 3D CC models reconstructed based on the segmentations of cross-sections of MR images were also visualised. The results, as measured by quantitative tests of the similarity indice (SI) and overlap value (OV) are the best for the EM model (SI = 92%, OV = 82%) and are comparable to or better than those for other methods taken from a literature review. Furthermore, the properties of the EM model consisting in its ability to both expand and shrink at the same time allow segmentations to be better fitted in subsequent CC slices then in state-of-the art ACMs such as DRLSE or SBGFRLS. The CC contours from previous and subsequent iterations produced by the EM model can be used for initiation in subsequent or previous frames of MR images, which makes the segmentation process easier, particularly as the CC area can increase or decrease in subsequent MR image frames.

Segmentation of lung fields from chest radiographs-a radiomic feature-based approach.

Precisely segmented lung fields restrict the region-of-interest from which radiological patterns are searched, and is thus an indispensable prerequisite step in any chest radiographic CADx system. Recently, a number of deep learning-based approaches have been proposed to implement this step. However, deep learning has its own limitations and cannot be used in resource-constrained settings. Medical systems generally have limited RAM, computational power, storage, and no GPUs. They are thus not always suited for running deep learning-based models. Shallow learning-based models with appropriately selected features give comparable performance but with modest resources. The present paper thus proposes a shallow learning-based method that makes use of 40 radiomic features to segment lung fields from chest radiographs. A distance regularized level set evolution (DRLSE) method along with other post-processing steps are used to refine its output. The proposed method is trained and tested using publicly available JSRT dataset. The testing results indicate that the performance of the proposed method is comparable to the state-of-the-art deep learning-based lung field segmentation (LFS) methods and better than other LFS methods.

Level set evolution driven by optimized area energy term for image segmentation

As a classic and famous active contour model for image segmentation, distance regularized level set evolution method avoids the process of re-initialization and can segment images flexibly, but it is easy to leak from objects with weak boundaries and fall into false boundaries. In this paper, an improved level set evolution model is proposed, in which an optimized area energy term combining a region growing matrix and an adaptive boundary indicator function is added to effectively detect boundaries for images with several adjacent targets and accelerate convergence at the same time. With an adaptive boundary indicator function involving a threshold defined by the standard deviation of images to be detected, this model can cross false boundaries and implement a correct segmentation for low contrast images. Meanwhile, the double-well potential function is optimized to make the model more stable. Experimental results on images of different objects have proved that the proposed model not only improves the precision of locating boundaries but also reduces the computational cost and works a stronger robustness than some other edge-based active contour models.

A deep Boltzmann machine-driven level set method for heart motion tracking using cine MRI images.

Heart motion tracking for radiation therapy treatment planning can result in effective motion management strategies to minimize radiation-induced cardiotoxicity. However, automatic heart motion tracking is challenging due to factors that include the complex spatial relationship between the heart and its neighboring structures, dynamic changes in heart shape, and limited image contrast, resolution, and volume coverage. In this study, we developed and evaluated a deep generative shape model-driven level set method to address these challenges. The proposed heart motion tracking method makes use of a heart shape model that characterizes the statistical variations in heart shapes present in a training data set. This heart shape model was established by training a three-layered deep Boltzmann machine (DBM) in order to characterize both local and global heart shape variations. During the tracking phase, a distance regularized level-set evolution (DRLSE) method was applied to delineate the heart contour on each frame of a cine MRI image sequence. The trained shape model was embedded into the DRLSE method as a shape prior term to constrain an evolutional shape to reach the desired heart boundary. Frame-by-frame heart motion tracking was achieved by iteratively mapping the obtained heart contour for each frame to the next frame as a reliable initialization, and performing a level-set evolution. The performance of the proposed motion tracking method was demonstrated using thirty-eight coronal cine MRI image sequences.

Mass Segmentation in Automated 3-D Breast Ultrasound Using Adaptive Region Growing and Supervised Edge-Based Deformable Model.

Automated 3-D breast ultrasound has been proposed as a complementary modality to mammography for early detection of breast cancers. To facilitate the interpretation of these images, computer aided detection systems are being developed in which mass segmentation is an essential component for feature extraction and temporal comparisons. However, automated segmentation of masses is challenging because of the large variety in shape, size, and texture of these 3-D objects. In this paper, the authors aim to develop a computerized segmentation system, which uses a seed position as the only priori of the problem. A two-stage segmentation approach has been proposed incorporating shape information of training masses. At the first stage, a new adaptive region growing algorithm is used to give a rough estimation of the mass boundary. The similarity threshold of the proposed algorithm is determined using a Gaussian mixture model based on the volume and circularity of the training masses. In the second stage, a novel geometric edge-based deformable model is introduced using the result of the first stage as the initial contour. In a data set of 50 masses, including 38 malignant and 12 benign lesions, the proposed segmentation method achieved a mean Dice of 0.74 ± 0.19 which outperformed the adaptive region growing with a mean Dice of 0.65 ± 0.2 (p-value < 0.02). Moreover, the resulting mean Dice was significantly (p-value < 0.001) better than that of the distance regularized level set evolution method (0.52 ± 0.27). The supervised method presented in this paper achieved accurate mass segmentation results in terms of Dice measure. The suggested segmentation method can be utilized in two aspects: 1) to automatically measure the change in volume of breast lesions over time and 2) to extract features for a computer aided detection or diagnosis system.

Spatially constrained distance regularized level set evolution method for segmentation of hydrops fetalis from ultrasound fetal heart images

Hydrops Fetalis is an abnormal condition in the fetus by an accretion of unusual fluid or edema in two or more fetal cavities for instance ascites, pericardial and pleural effusion. Accumulation of atypical fluid above 2 mm needs extra attention of radiologists. Ultrasound imaging technique helps to diagnose the fluid accumulation within the pericardial tissues. The trained medicinal practitioners also find difficult to analyse the hydrops due to the low assessment gray scale ultrasound images. The segmentation of hydrops fetalis from the fetal heart ultrasound image helps the sonographists and pediatrists to diagnose easily and effectively at the early stage itself. The low quality and poor resolution ultrasound image, make the segmentation process difficult. We use the spatially constrained-distance regularized level set evolution algorithm for segmenting the hydrops region.

An Improved Adaptive Level Set Method for Image Segmentation

In order to improve the accuracy of image segmentation, an improved adaptive level set method is proposed based on level set evolution without re-initialization method and adaptive distance preserving level set evolution method. A new definition of weight coefficient in evolution equations is the main innovation of this paper. The improved method can detect certain object boundaries, interior and exterior contours of an object, edges of multi-objects and weak boundaries of an object by synthetic and real images numerical experiments. Numerical results show that the improved adaptive level set method has faster segmentation speed and higher segmentation accuracy compared with the previous two methods, especially in weak boundaries and edges of multi-objects segmentation problems.

Topology optimization of laminated composite structures with design-dependent loads

A solution for the topology structures of composite laminates with design-dependent loads is presented. The methodology is based on the isoline method, the sensitivity filter with density gradient weighting, and the solid isotropic material with penalization (SIMP) method. In addition, the element stiffness matrix for composite layups is derived. Two- and three-joint constraint examples of the topology optimization design for composite structures are presented. The work efficiency of the isoline method and the distance regularized level set evolution (DRLSE) method are compared; five other filtering methods are also studied. The distribution rule of solid material and the relationship between the minimum compliance of optimal structures and the principal angle of composite layups are discussed.

Level Set Based Hippocampus Segmentation in MR Images with Improved Initialization Using Region Growing.

The hippocampus has been known as one of the most important structures referred to as Alzheimer's disease and other neurological disorders. However, segmentation of the hippocampus from MR images is still a challenging task due to its small size, complex shape, low contrast, and discontinuous boundaries. For the accurate and efficient detection of the hippocampus, a new image segmentation method based on adaptive region growing and level set algorithm is proposed. Firstly, adaptive region growing and morphological operations are performed in the target regions and its output is used for the initial contour of level set evolution method. Then, an improved edge-based level set method utilizing global Gaussian distributions with different means and variances is developed to implement the accurate segmentation. Finally, gradient descent method is adopted to get the minimization of the energy equation. As proved by experiment results, the proposed method can ideally extract the contours of the hippocampus that are very close to manual segmentation drawn by specialists.

Medical Imaging Lesion Detection Based on Unified Gravitational Fuzzy Clustering

We develop a swift, robust, and practical tool for detecting brain lesions with minimal user intervention to assist clinicians and researchers in the diagnosis process, radiosurgery planning, and assessment of the patient's response to the therapy. We propose a unified gravitational fuzzy clustering-based segmentation algorithm, which integrates the Newtonian concept of gravity into fuzzy clustering. We first perform fuzzy rule-based image enhancement on our database which is comprised of T1/T2 weighted magnetic resonance (MR) and fluid-attenuated inversion recovery (FLAIR) images to facilitate a smoother segmentation. The scalar output obtained is fed into a gravitational fuzzy clustering algorithm, which separates healthy structures from the unhealthy. Finally, the lesion contour is automatically outlined through the initialization-free level set evolution method. An advantage of this lesion detection algorithm is its precision and its simultaneous use of features computed from the intensity properties of the MR scan in a cascading pattern, which makes the computation fast, robust, and self-contained. Furthermore, we validate our algorithm with large-scale experiments using clinical and synthetic brain lesion datasets. As a result, an 84%–93% overlap performance is obtained, with an emphasis on robustness with respect to different and heterogeneous types of lesion and a swift computation time.

Structural topology optimization with design-dependent pressure loads

One way to solve topology optimization of continuum structures under design-dependent pressure loads is to recover the loading surface at each step of the minimization process. During the topology evolution, the intermediate topologies obtained by using the SIMP (Solid Isotropic Material with Penalization) method actually can be regarded as gray scale images, for which the paper proposes a new material boundary identification scheme based on image segmentation technique. The Distance Regularized Level Set Evolution (DRLSE) method proposed by Li et al., IEEE Trans Image Process 19(12):3243---3254 (2010) is utilized to detect the image edge. Then the pressure boundary is represented as the zero level contour of a level set function (LSF). Inheriting the merits of the level set method, the current scheme can handle the topological change of the pressure boundary efficiently and be easily extended to the three-dimensional problems. In addition, the scheme is more stable compared with the conventional loading surface searching methods since it works well for the intermediate topologies with local scattered densities. A new optimization framework is also proposed to avoid the load sensitivity analysis. Four numerical examples are presented to show the validity and advantages of the proposed scheme.

A new active contour remote sensing river image segmentation algorithm inspired from the cross entropy

The CV (Chan–Vese) model is a piecewise constant approximation of the Mumford and Shah model. It assumes that the original image can be segmented into two regions such that each region can be represented as constant grayscale value. In fact, the objective functional of the CV model actually finds a segmentation of the image such that the within-class variance is minimized. This is equivalent to the Otsu image thresholding algorithm which also aims to minimize the within-class variance. Similarly to the Otsu image thresholding algorithm, cross entropy is another widely used image thresholding algorithm and it finds a segmentation such that the cross entropy of the segmented image and the original image is minimized. Inspired from the cross entropy, a new active contour image segmentation algorithm is proposed. The region term in the new objective functional is the integral of the logarithm of the ratio between the grayscale of the original image and the mean value computed from the segmented image weighted by the grayscale of the original image. The new objective functional can be solved by the level set evolution method. A distance regularized term is added to the level set evolution equation so the level set need not be reinitialized periodically. A fast global minimization algorithm of the objective functional is also proposed which incorporates the edge term originated from the geodesic active contour model. Experimental results show that, the algorithm proposed can segment images more accurately than the CV model and the implementation speed of the fast global minimization algorithm is fast.

Heuristic Initialization for Active Contour Models in CT/MRI Image Processing

It is an important segmentation approach of CT/MRI images to automatically extract contours in every slice using active contour models. The key point of the segmentation approach is to automatically construct initial contours for active contour models because any active contour model is sensitive to its initial contour. This paper presents an algorithm to construct such initial contours using a heuristic method. Assume that the contour in previous slice (previous contour) is accurate. The contour in the current slice (current contour) is constructed according to the previous contour using the way: Recognition and link of edge points of tissues according to the previous contour. The contour linking edge points is used as the initial contour of the distance regularized level set evolution (DRLSE) method and then an accurate contour can be extracted in the current slice.

Comparison of Two Iris Localization Algorithms

Iris recognition is regarded as a most reliable and accurate biometric identification system. Daugman’s Integrodifferential operator is a linear search method which makes the identification process extremely slow as well as increases the false acceptance rate beyond an acceptable range. The present work uses distance regularized level set evolution (DRLSE) method on CASIA-V3-Interval database and applies a suitable algorithm to detect the iris from an image. The two techniques i.e., Daugman’s Integro-differential operator and DRLSE are compared based on accuracy and time taken to localize the iris.

Segmentation of renal parenchymal area from ultrasound images using level set evolution.

This paper presents a framework for segmentation of renal parenchymal area from ultrasound images based on a 2-step level set method. We used distance regularized level set evolution method to partition the kidney boundary, followed by region-scalable fitting energy minimization method to segment the kidney collecting system, and determined renal parenchymal area by subtracting the area of the collecting system from the gross kidney area. The proposed method demonstrated excellent validity and low inter-observer variability.