医学图像中典型分割算法的比较分析

通过图像分割算法获取医学图像感兴趣区域是医学图像分析与识别要解决的首要问题及技术难点。在阐述医学图像分割及其研究意义的基础上,总结了医学图像分割中的典型算法。接着,以MR颅脑图像的分割为例,做了基于FCM、C-V、区域生长法的图像分割及对比分析,讨论了医学图像有效分割存在的技术难点。最后,指明了医学图像分割未来的发展趋势。研究成果对深入认识医学图像分割的研究意义、分割算法、分割难点、分割趋势等,推动医学图像分割技术的发展均具有一定意义。     

Semi-supervised superpixel classification for medical images segmentation: application to detection of glaucoma disease

Glaucoma is a disease characterized by damaging the optic nerve head, this can result in severe vision loss. An early detection and a good treatment provided by the ophthalmologist are the keys to preventing optic nerve damage and vision loss from glaucoma. Its screening is based on the manual optic cup and disc segmentation to measure the vertical cup to disc ratio (CDR). However, obtaining the regions of interest by the expert ophthalmologist can be difficult and is often a tedious task. In most cases, the unlabeled images are more numerous than the labeled ones.We propose an automatic glaucoma screening approach named Super Pixels for Semi-Supervised Segmentation “SP3S”, which is a semi-supervised superpixel-by-superpixel classification method, consisting of three main steps. The first step has to prepare the labeled and unlabeled data, applying the superpixel method and bringing in an expert for the labeling of superpixels. In the second step, We incorporate prior knowledge of the optic cup and disc by including color and spatial information. In the final step, semi-supervised learning by the Co-forest classifier is trained only with a few number of labeled superpixels and a large number of unlabeled superpixels to generate a robust classifier. For the estimation of the optic cup and disc regions, the active geometric shape model is used to smooth the disc and cup boundary for the calculation of the CDR. The obtained results for glaucoma detection, via an automatic cup and disc segmentation, established a potential solution for glaucoma screening. The SP3S performance shows quantitatively and qualitatively similar correspondence with the expert segmentation, providing an interesting tool for semi-automatic recognition of the optic cup and disc in order to achieve a medical progress of glaucoma disease.     

Tooth segmentation of dental study models using range images

The accurate segmentation of the teeth from the digitized representation of a dental study model is an important component in computer-based algorithms for orthodontic feature detection and measurement and in the simulation of orthodontic procedures such as tooth rearrangement. This paper presents an automated method for tooth segmentation from the three-dimensional (3-D) digitized image captured by a laser scanner. We avoid the complexity of directly processing 3-D mesh data by proposing the innovative idea of detecting features on two range images computed from the 3-D image. The dental arch is first obtained from the plan-view range image. Using the arch as the reference, a panoramic range image of the dental model can be computed. The interstices between the teeth are detected separately in the two range images, and results from both views are combined for a determination of interstice locations and orientations. Finally, the teeth are separated from the gums by delineating the gum margin. The algorithm was tested on 34 dental models representing a variety of malocclusions and was found to be robust and accurate.     

A variational method for geometric regularization of vascular segmentation in medical images

In this paper, a level-set-based geometric regularization method is proposed which has the ability to estimate the local orientation of the evolving front and utilize it as shape induced information for anisotropic propagation. We show that preserving anisotropic fronts can improve elongations of the extracted structures, while minimizing the risk of leakage. To that end, for an evolving front using its shape-offset level-set representation, a novel energy functional is defined. It is shown that constrained optimization of this functional results in an anisotropic expansion flow which is usefull for vessel segmentation. We have validated our method using synthetic data sets, 2-D retinal angiogram images and magnetic resonance angiography volumetric data sets. A comparison has been made with two state-of-the-art vessel segmentation methods. Quantitative results, as well as qualitative comparisons of segmentations, indicate that our regularization method is a promising tool to improve the efficiency of both techniques.     

Knowledge-based segmentation of Landsat images

A knowledge-based approach for Landsat image segmentation is proposed. The image segmentation problem is solved by extracting kernel information from the input image to provide an initial interpretation of the image and by using a knowledge-based hierarchical classifier to discriminate between major land-cover types in the study area. The proposed method is designed in such a way that a Landsat image can be segmented and interpreted without any prior image-dependent information. The general spectral land-cover knowledge is constructed from the training land-cover data, and the road information of an image is obtained through a road-detection program     

Intensity-based segmentation of microarray images

The underlying principle in microarray image analysis is that the spot intensity is a measure of the gene expression. This implicitly assumes the gene expression of a spot to be governed entirely by the distribution of the pixel intensities. Thus, a segmentation technique based on the distribution of the pixel intensities is appropriate for the current problem. In this paper, clustering-based segmentation is described to extract the target intensity of the spots. The approximate boundaries of the spots in the microarray are determined by manual adjustment of rectilinear grids. The distribution of the pixel intensity in a grid containing a spot is assumed to be the superposition of the foreground and the local background. The k-means clustering technique and the partitioning around medoids (PAM) were used to generate a binary partition of the pixel intensity distribution. The median (k-means) and the medoid (PAM) of the cluster members are chosen as the cluster representatives. The effectiveness of the clustering-based segmentation techniques was tested on publicly available arrays generated in a lipid metabolism experiment (Callow et al., 2000). The results are compared against those obtained using the region-growing approach (SPOT) (Yang et al., 2001). The effect of additive white Gaussian noise is also investigated.     

Optimal segmentation of cell images

An optimal segmentation algorithm for light microscopic cell images is presented. The image segmentation is performed by thresholding a parametric image approximating the original image. Using the mean squared error between the original and the constructed image as the cost function, the segmentation problem is transformed into an optimisation process where parametric parameters are determined that minimise the defined cost function. The cost function is iteratively minimised using an unsupervised learning rule to adjust the parameters, and a parametric image is constructed at each iteration, based on the obtained parameters. The cell region is extracted by thresholding the final parametric image, where the threshold is one of the image parameters. Application results to real cervical images are provided to show the performance of the proposed segmentation approach. Experimental segmentation results are presented for the proposed optimal algorithm for synthetic cell images corrupted by variant levels of noise; these results are compared with the K-means clustering method and Bayes classifier in terms of classification errors     

Unsupervised segmentation of textured images using a hierarchicalneural structure

A hierarchical learning structure, combining a randomly-placed local window, a self-organising map and a local-voting scheme, has been developed for the unsupervised segmentation of textured images, which are modelled by Markov random fields. The system learns to progressively estimate model parameters, and hence classify the various textured regions. A globally correct segregation has consistently been obtained during extensive experiments on both synthetic and natural textured images     

Rule-based automatic segmentation of color images

A rule-based segmentation algorithm for color images has been presented in this paper. The proposed strategy is similar to region growing algorithm where the seed points are automatically selected and grown. The similarity percents of neighboring pixels are calculated by means of fuzzy reasoning rules, and the merging of the pixels with regions is performed by comparing the similarity percent with the similarity threshold value. The algorithm does not require any prior knowledge of the number of regions existing in the image and decreases the computational load required for the fuzzy c-means (FCM). Several computer simulations have been performed and the results have been discussed. The simulation results indicate that the proposed algorithm yields segmented color image of perfect accuracy.     

Multiscale morphological segmentation of gray-scale images

In this paper, the authors have proposed a method of segmenting gray level images using multiscale morphology. The approach resembles the watershed algorithm in the sense that the dark (respectively bright) features which are basically canyons (respectively mountains) on the surface topography of the gray level image are gradually filled (respectively clipped) using multiscale morphological closing (respectively opening) by reconstruction with isotropic structuring element. The algorithm detects valid segments at each scale using three criteria namely growing, merging and saturation. Segments extracted at various scales are integrated in the final result. The algorithm is composed of two passes preceded by a preprocessing step for simplifying small scale details of the image that might cause over-segmentation. In the first pass feature images at various scales are extracted and kept in respective level of morphological towers. In the second pass, potential features contributing to the formation of segments at various scales are detected. Finally the algorithm traces the contours of all such contributing features at various scales. The scheme after its implementation is executed on a set of test images (synthetic as well as real) and the results are compared with those of few other standard methods. A quantitative measure of performance is also formulated for comparing the methods.     

Precise segmentation of multimodal images.

We propose new techniques for unsupervised segmentation of multimodal grayscale images such that each region-of-interest relates to a single dominant mode of the empirical marginal probability distribution of grey levels. We follow the most conventional approaches in that initial images and desired maps of regions are described by a joint Markov-Gibbs random field (MGRF) model of independent image signals and interdependent region labels. However, our focus is on more accurate model identification. To better specify region borders, each empirical distribution of image signals is precisely approximated by a linear combination of Gaussians (LCG) with positive and negative components. We modify an expectation-maximization (EM) algorithm to deal with the LCGs and also propose a novel EM-based sequential technique to get a close initial LCG approximation with which the modified EM algorithm should start. The proposed technique identifies individual LCG models in a mixed empirical distribution, including the number of positive and negative Gaussians. Initial segmentation based on the LCG models is then iteratively refined by using the MGRF with analytically estimated potentials. The convergence of the overall segmentation algorithm at each stage is discussed. Experiments show that the developed techniques segment different types of complex multimodal medical images more accurately than other known algorithms.     

Adaptive fuzzy segmentation of magnetic resonance images

An algorithm is presented for the fuzzy segmentation of two-dimensional (2-D) and three-dimensional (3-D) multispectral magnetic resonance (MR) images that have been corrupted by intensity inhomogeneities, also known as shading artifacts. The algorithm is an extension of the 2-D adaptive fuzzy C-means algorithm (2-D AFCM) presented in previous work by the authors. This algorithm models the intensity inhomogeneities as a gain field that causes image intensities to smoothly and slowly vary through the image space. It iteratively adapts to the intensity inhomogeneities and is completely automated. In this paper, we fully generalize 2-D AFCM to three-dimensional (3-D) multispectral images. Because of the potential size of 3-D image data, we also describe a new faster multigrid-based algorithm for its implementation. We show, using simulated MR data, that 3-D AFCM yields lower error rates than both the standard fuzzy C-means (FCM) algorithm and two other competing methods, when segmenting corrupted images. Its efficacy is further demonstrated using real 3-D scalar and multispectral MR brain images.     

SEM algorithm and unsupervised statistical segmentation ofsatellite images

The work addresses Bayesian unsupervised satellite image segmentation, using contextual methods. It is shown, via a simulation study, that the spatial or spectral context contribution is sensitive to image parameters such as homogeneity, means, variances, and spatial or spectral correlations of the noise. From this one may choose the best context contribution according to the estimated values of the above parameters. The parameter estimation is done by SEM, a densities mixture estimator which is a stochastic variant of the EM (expectation-maximization) algorithm. Another simulation study shows good robustness of the SEM algorithm with respect to different image parameters. Thus, modification of the behavior of the contextual methods, when the SEM-based unsupervised approaches are considered, is limited, and the conclusions of the supervised simulation study stay valid. An adaptive unsupervised method using more relevant contextual features is proposed. Different SEM-based unsupervised contextual segmentation methods, applied to two real SPOT images, give consistently better results than a classical histogram-based method     

Automated qualitative assessment of multi-modal distortions in digital images based on GLZ

This paper introduces a novel approach to a qualitative assessment of images affected by multi-modal distortions. The idea is to assess the image quality perceived by an end user in an automatic way in order to avoid the usual time-consuming, costly and non-repeatable method of collecting subjective scores during a psycho-physical experiment. This is achieved by computing quantitative image distortions and mapping results on qualitative scores. Useful mapping models have been proposed and constructed using the generalised linear model (GLZ), which is a generalisation of the least squares regression in statistics for ordinal data. Overall qualitative image distortion is computed based on partial quantitative distortions from component algorithms operating on specified image features. Seven such algorithms are applied to successfully analyse the seven image distortions in relation to the original image. A survey of over 12,000 subjective quality scores has been carried out in order to determine the influence of these features on the perceived image quality. The results of quantitative assessments are mapped on the surveyed scores to obtain an overall quality score of the image. The proposed models have been validated in order to prove that the above technique can be applied to automatic image quality assessment.     

Two-stage joint non-causal filtering and segmentation of nonuniform images

The optimal and quasioptimal algorithms of two-stage non-causal filtering and segmentation of nonuniform images distorted by additive interference with independent values of samples have been synthesized on the basis of the mathematical tools of mixed Markov discrete time processes. The first stage involves the performance of one-dimensional joint filtering and segmentation of nonuniform images along rows and columns. The second stage involves the combining of estimates computed at the first stage. The analysis of the quasioptimal algorithm was performed by using the computer statistical simulation of model example.     

Image segmentation for compression of images and image sequences

Region-based image coding schemes, the so-called second generation techniques, have gained much favour in recent years. For still picture coding, they can increase the compression ratio obtained by transform coding by an order of magnitude, while maintaining adequate image representation. The success of these techniques relies on the ability to describe regions in an image succinctly by their shape and size. The algorithms presented describe methods for segmenting images. Unlike most other region based algorithms, these algorithms incorporate knowledge of the border coding process in deciding how to partition the image. The extension from single image compression to sequential image compression is also considered. A new, efficient segmentation scheme is proposed that exploits temporal redundancies between successive images, and reduces some problems associated with error accumulation in error images     

Two-stage mutual causal filtration and segmentation of heterogeneous images

Using the mathematical technique of mixed Markovian processes in discrete time optimal and quasi-optimal algorithms that combine results of one-dimensional filtration and segmentation of heterogeneous images are synthesized. Analysis of the quasi-optimal algorithm is conducted on a model example using statistical modeling on PC.     

Three-dimensional segmentation and interpolation of magnetic resonance brain images

The authors propose a method for the 3-D reconstruction of the brain from anisotropic magnetic resonance imaging (MRI) brain data. The method essentially consists in two original algorithms both for segmentation and for interpolation of the MRI data. The segmentation process is performed in three steps. A gray level thresholding of the white and gray matter tissue is performed on the brain MR raw data. A global white matter segmentation is automatically performed with a global 3-D connectivity algorithm which takes into account the anisotropy of the MRI voxel. The gray matter is segmented with a local 3-D connectivity algorithm. Mathematical morphology tools are used to interpolate slices. The whole process gives an isotropic binary representation of both gray and white matter which are available for 3-D surface rendering. The power and practicality of this method have been tested on four brain datasets. The segmentation algorithm favorably compares to a manual one. The interpolation algorithm was compared to the shaped-based method both quantitatively and qualitatively.