SEARCH · PubMed Health

Results for “Automated image segmentation”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 19 recordsLinked to original sources

PubMed2000

Object localization and border detection criteria design in edge-based image segmentation: automated learning from examples.

This paper provides methodology for fully automated model-based image segmentation. All information necessary to perform image segmentation is automatically derived from a training set that is presented in a form of segmentation examples. The training set is used to construct two models representing the objects--shape model and border appearance model. A two-step approach to image segmentation is reported. In the first step, an approximate location of the object of interest is determined. In the second step, accurate border segmentation is performed. The shape-variant Hough transform method was developed that provides robust object localization automatically. It finds objects of arbitrary shape, rotation, or scaling and can handle object variability. The border appearance model was developed to automatically design cost functions that can be used in the segmentation criteria of edge-based segmentation methods. Our method was tested in five different segmentation tasks that included 489 objects to be segmented. The final segmentation was compared to manually defined borders with good results [rms errors in pixels: 1.2 (cerebellum), 1.1 (corpus callosum), 1.5 (vertebrae), 1.4 (epicardial), and 1.6 (endocardial) borders]. Two major problems of the state-of-the-art edge-based image segmentation algorithms were addressed: strong dependency on a close-to-target initialization, and necessity for manual redesign of segmentation criteria whenever new segmentation problem is encountered.

Algorithms↗

PubMed1990

Automated image segmentation and serial section reconstruction in microscopy.

A technique for automatic 3-D reconstruction of specifically stained features in televised serial histological sections has been developed using an image analyser. Images which included these features were individually converted to binary images, compiled, displayed to show the 3-D morphology and used to quantify the structure. The criteria necessary for producing valid reconstructions and the problems associated with the manipulation of images of fine detail, particularly those containing several thousand features, are illustrated by examples from skin.

Animals↗

PubMed1989

Automated image segmentation by fractal grey tone functions.

A new image segmentation method is presented. It is based on mathematical derivations stemming from both mathematical morphology and fractal geometry. Structures of different nature are most often characterized by a fractal grey tone function, in a certain range of resolutions, with a particular value of the fractal dimension. Our method allows the isolation of such structures. It only requires a series of dilated and eroded grey tone images.

Algorithms↗

PubMedDate not supplied

Selection of MR images for automated segmentation.

MR images show a large range of contrast for various tissues in the body and are ideal for multispectral segmentation. Typically, only two MR images (dual-echo series) are used for segmentation; however, other images are often available. We evaluated MR images from 40 patients to determine the optimal type and number of images required for segmentation of tissues associated with brain tumors (normal brain, edema, necrosis, and active tumor). Pattern recognition methods indicated that three MR images from the same slice location were adequate for segmentation, as defined by feature selection and feature extraction measures based on training fields. This result was also confirmed by visually examining segmented images for all 40 patients. This work demonstrates that by using existing image/statistical analysis techniques (feature selection and feature extraction), one can systematically determine the optimal type and number of MR images for tissue segmentation.

Brain Neoplasms↗

PubMedDate not supplied

Fully automated biomedical image segmentation by self-organized model adaptation.

In this paper, we present a fully automated image segmentation method based on an algorithm that provides adaptive plasticity in function approximation problems: the deformable (feature) map (DM) algorithm. The DM approach reduces a class of similar function approximation problems to the explicit supervised one-shot training of a single data set. This is followed by a subsequent, appropriate similarity transformation, which is based on a self-organized deformation of the underlying multidimensional probability distributions. We apply this algorithm to the real-world problem of fully automated voxel-based multispectral image segmentation, employing magnetic resonance data sets of the human brain. In contrast to previous segmentation approaches, the knowledge obtained within the segmentation procedure of a single prototypical reference data set can be re-utilized for the segmentation of new, 'similar' data employing a strategy of incremental adaptive learning based on the DM algorithm. Thus, we obtain a fully automatic segmentation method that does neither require manual contour tracing of training regions, visual classification of voxel clusters, nor any other kind of human intervention. Our application demonstrates that flexible learning by a strategy of self-organized incremental model adaptation can contribute to increase the efficiency and practicability of biomedical image processing systems.

Algorithms↗

PubMed1998

Advances in high-speed, three-dimensional imaging and automated segmentation algorithms for thick and overlapped clusters in cytologic preparations. Application to cervical smears.

OBJECTIVE: To use three-dimensional (3-D) imaging and localized adaptive image analysis to enable automated cervical smear screening systems to efficiently and effectively process thick and overlapped cell clusters currently left unprocessed. STUDY DESIGN: Instrumentation was developed to perform high-speed (50-200 optical sections per second at 256 x 256 resolution), 3-D imaging of thick regions of cervical smears. Normal and abnormal ThinPrep smears were imaged at two levels of resolution to approximate higher-resolution, wide-area imaging. Improved dual-resolution, 3-D image analysis algorithms were developed for segmenting nuclei in these clusters. RESULTS: Despite low contrast, high variability and dense overlaps, the algorithms detected 89% and correctly segmented 76% of nuclei in clusters from normal smears and detected 75% and correctly segmented 45% of nuclei in clusters from abnormal smears in low-resolution images. In high-resolution images they detected 88% and segmented 76% of nuclei from normal specimens and detected 55% and segmented 45% of nuclei from abnormal specimens. At least one nucleus from each cell cluster was correctly segmented. CONCLUSION: Selective application of 3-D imaging and 3-D image analysis to thick and overlapped regions can enable a significant fraction (45-89%) of clustered and embedded cells to be accessed by an automated analysis system. These regions are, for the most part, unprocessable by current two-dimensional methods.

Algorithms↗

PubMed2001

Automated segmentation of MR images of brain tumors.

An automated brain tumor segmentation method was developed and validated against manual segmentation with three-dimensional magnetic resonance images in 20 patients with meningiomas and low-grade gliomas. The automated method (operator time, 5-10 minutes) allowed rapid identification of brain and tumor tissue with an accuracy and reproducibility comparable to those of manual segmentation (operator time, 3-5 hours), making automated segmentation practical for low-grade gliomas and meningiomas.

Brain↗

PubMed2004

Three validation metrics for automated probabilistic image segmentation of brain tumours.

The validity of brain tumour segmentation is an important issue in image processing because it has a direct impact on surgical planning. We examined the segmentation accuracy based on three two-sample validation metrics against the estimated composite latent gold standard, which was derived from several experts' manual segmentations by an EM algorithm. The distribution functions of the tumour and control pixel data were parametrically assumed to be a mixture of two beta distributions with different shape parameters. We estimated the corresponding receiver operating characteristic curve, Dice similarity coefficient, and mutual information, over all possible decision thresholds. Based on each validation metric, an optimal threshold was then computed via maximization. We illustrated these methods on MR imaging data from nine brain tumour cases of three different tumour types, each consisting of a large number of pixels. The automated segmentation yielded satisfactory accuracy with varied optimal thresholds. The performances of these validation metrics were also investigated via Monte Carlo simulation. Extensions of incorporating spatial correlation structures using a Markov random field model were considered.

Algorithms↗

PubMed1995

Knowledge-based medical image analysis and representation for integrating content definition with the radiological report.

Technology breakthroughs in high-speed, high-capacity, and high performance desk-top computers and workstations make the possibility of integrating multimedia medical data to better support clinical decision making, computer-aided education, and research not only attractive, but feasible. To systematically evaluate results from increasingly automated image segmentation it is necessary to correlate them with the expert judgments of radiologists and other clinical specialists interpreting the images. These are contained in increasingly computerized radiological reports and other related clinical records. But to make automated comparison feasible it is necessary to first ensure compatibility of the knowledge content of images with the descriptions contained in these records. Enough common vocabulary, language, and knowledge representation components must be represented on the computer, followed by automated extraction of image-content descriptions from the text, which can then be matched to the results of automated image segmentation. A knowledge-based approach to image segmentation is essential to obtain the structured image descriptions needed for matching against the expert's descriptions. We have developed a new approach to medical image analysis which helps generate such descriptions: a knowledge-based object-centered hierarchical planning method for automatically composing the image analysis processes. The problem-solving steps of specialists are represented at the knowledge level in terms of goals, tasks, and domain objects and concepts separately from the implementation level for specific representations of different image types, and generic analysis methods. This system can serve as a major functional component in incrementally building and updating a structured and integrated hybrid information system of patient data.(ABSTRACT TRUNCATED AT 250 WORDS)

Artificial Intelligence↗

PubMed1998

Automated segmentation of cell nuclei in fine needle aspirates of the breast.

OBJECTIVE: To achieve a classifier of breast lesions to discriminate between benign and malignant cases of cytologic smears with automated segmentation image analysis techniques. STUDY DESIGN: The techniques were applied to images of epithelial cell nuclei from cytologic smears obtained by fine needle aspiration. The images of the nuclei were taken from 95 cases of malignant lesions and 47 benign (approximately 25 nuclei per case), and 28 nuclear variables were measured. The data were analyzed by a double methodology, discriminant analysis and classification and regression trees (CART), to determine which provided the best results. RESULTS: CART selected the SD of the nuclear area with correct classification of 85.1% of benign and 94.7% malignant aspirates. Discriminant analysis selected the group of variables formed by axis lengths, SD of the longest axis, sphericity and variance of gray levels, with results similar to those of CART. CONCLUSION: Automated segmentation image analysis techniques were effective, and the classifier was quick, simple and efficacious in malignant-benign discrimination.

Algorithms↗

PubMed2002

Adaptive elastic segmentation of brain MRI via shape-model-guided evolutionary programming.

This paper presents a fully automated segmentation method for medical images. The goal is to localize and parameterize a variety of types of structure in these images for subsequent quantitative analysis. We propose a new hybrid strategy that combines a general elastic template matching approach and an evolutionary heuristic. The evolutionary algorithm uses prior statistical information about the shape of the target structure to control the behavior of a number of deformable templates. Each template, modeled in the form of a B-spline, is warped in a potential field which is itself dynamically adapted. Such a hybrid scheme proves to be promising: by maintaining a population of templates, we cover a large domain of the solution space under the global guidance of the evolutionary heuristic, and thoroughly explore interesting areas. We address key issues of automated image segmentation systems. The potential fields are initially designed based on the spatial features of the edges in the input image, and are subjected to spatially adaptive diffusion to guarantee the deformation of the template. This also improves its global consistency and convergence speed. The deformation algorithm can modify the internal structure of the templates to allow a better match. We investigate in detail the preprocessing phase that the images undergo before they can be used more effectively in the iterative elastic matching procedure: a texture classifier, trained via linear discriminant analysis of a learning set, is used to enhance the contrast of the target structure with respect to surrounding tissues. We show how these techniques interact within a statistically driven evolutionary scheme to achieve a better tradeoff between template flexibility and sensitivity to noise and outliers. We focus on understanding the features of template matching that are most beneficial in terms of the achieved match. Examples from simulated and real image data are discussed, with considerations of algorithmic efficiency.

Algorithms↗

PubMed2005

Towards automated cellular image segmentation for RNAi genome-wide screening.

The Rho family of small GTPases is essential for morphological changes during normal cell development and migration, as well as during disease states such as cancer. Our goal is to identify novel effectors of Rho proteins using a cell-based assay for Rho activity to perform genome-wide functional screens using double stranded RNA (dsRNAs) interference. We aim to discover genes could cause the cell phenotype changed dramatically. Biologists currently attempt to perform the genome-wide RNAi screening to identify various image phenotypes. RNAi genome-wide screening, however, could easily generate more than a million of images per study, manual analysis is thus prohibitive. Image analysis becomes a bottleneck in realizing high content imaging screens. We propose a two-step segmentation approach to solve this problem. First, we determine the center of a cell using the information in the DNA-channel by segmenting the DNA nuclei and the dissimilarity function is employed to attenuate the over-segmentation problem, then we estimate a rough boundary for each cell using a polygon. Second, we apply fuzzy c-means based multi-threshold segmentation and sharpening technology; for isolation of touching spots, marker-controlled watershed is employed to remove touching cells. Furthermore, Voronoi diagrams are employed to correct the segmentation errors caused by overlapping cells. Image features are extracted for each cell. K-nearest neighbor classifier (KNN) is employed to perform cell phenotype classification. Experimental results indicate that the proposed approach can be used to identify cell phenotypes of RNAi genome-wide screens.

Algorithms↗

PubMed2003

Automated cardiac MR image segmentation: theory and measurement evaluation.

We present a new approach to magnetic resonance image segmentation with a Gradient-Vector-Flow-based snake applied to selective smoothing filtered images. The system also allows automated image segmentation in the presence of grey scale inhomogeneity, as in cardiac Magnetic Resonance imaging. Removal of such inhomogeneities is a difficult task, but we proved that using non-linear anisotropic diffusion filtering, myocardium edges are selectively preserved. The approach allowed medical data to be automatically segmented in order to track not only endocardium, which is usually a less difficult task, but also epicardium in anatomic and perfusion studies with Magnetic Resonance. The method developed proceeds in three distinct phases: (a) an anisotropic diffusion filtering tool is used to reduce grey scale inhomogeneity and to selectively preserve edges; (b) a Gradient-Vector-Flow-based snake is applied on filtered images to allow capturing a snake from a long range and to move into concave boundary regions; and (c) an automatic procedure based on a snake is used to fit both endocardium and epicardium borders in a multiphase, multislice examination. A good agreement (P<0.001) between manual and automatic data analysis, based on the mean difference+/-SD, was assessed in a pool of 907 cardiac function and perfusion images.

Algorithms↗

PubMed1997

Quantitative analysis of cardiovascular MR images.

The diagnosis of cardiovascular disease requires the precise assessment of both morphology and function. Nearly all aspects of cardiovascular function and flow can be quantified nowadays with fast magnetic resonance (MR) imaging techniques. Conventional and breath-hold cine MR imaging allow the precise and highly reproducible assessment of global and regional left ventricular function. During the same examination, velocity encoded cine (VEC) MR imaging provides measurements of blood flow in the heart and great vessels. Quantitative image analysis often still relies on manual tracing of contours in the images. Reliable automated or semi-automated image analysis software would be very helpful to overcome the limitations associated with the manual and tedious processing of the images. Recent progress in MR imaging of the coronary arteries and myocardial perfusion imaging with contrast media, along with the further development of faster imaging sequences, suggest that MR imaging could evolve into a single technique ('one stop shop') for the evaluation of many aspects of heart disease. As a result, it is very likely that the need for automated image segmentation and analysis software algorithms will further increase. In this paper the developments directed towards the automated image analysis and semi-automated contour detection for cardiovascular MR imaging are presented.

Blood Flow Velocity↗

PubMed2004

Simultaneous truth and performance level estimation (STAPLE): an algorithm for the validation of image segmentation.

Characterizing the performance of image segmentation approaches has been a persistent challenge. Performance analysis is important since segmentation algorithms often have limited accuracy and precision. Interactive drawing of the desired segmentation by human raters has often been the only acceptable approach, and yet suffers from intra-rater and inter-rater variability. Automated algorithms have been sought in order to remove the variability introduced by raters, but such algorithms must be assessed to ensure they are suitable for the task. The performance of raters (human or algorithmic) generating segmentations of medical images has been difficult to quantify because of the difficulty of obtaining or estimating a known true segmentation for clinical data. Although physical and digital phantoms can be constructed for which ground truth is known or readily estimated, such phantoms do not fully reflect clinical images due to the difficulty of constructing phantoms which reproduce the full range of imaging characteristics and normal and pathological anatomical variability observed in clinical data. Comparison to a collection of segmentations by raters is an attractive alternative since it can be carried out directly on the relevant clinical imaging data. However, the most appropriate measure or set of measures with which to compare such segmentations has not been clarified and several measures are used in practice. We present here an expectation-maximization algorithm for simultaneous truth and performance level estimation (STAPLE). The algorithm considers a collection of segmentations and computes a probabilistic estimate of the true segmentation and a measure of the performance level represented by each segmentation. The source of each segmentation in the collection may be an appropriately trained human rater or raters, or may be an automated segmentation algorithm. The probabilistic estimate of the true segmentation is formed by estimating an optimal combination of the segmentations, weighting each segmentation depending upon the estimated performance level, and incorporating a prior model for the spatial distribution of structures being segmented as well as spatial homogeneity constraints. STAPLE is straightforward to apply to clinical imaging data, it readily enables assessment of the performance of an automated image segmentation algorithm, and enables direct comparison of human rater and algorithm performance.

Algorithms↗

PubMed1996

Volume rendering of visible human data for an anatomical virtual environment.

In this work, we utilize the axial anatomical human male sections from the National Library of Medicine's Visible Human Project to generate three-dimensional (3-D) volume representations of the human male subject. The two-dimensional (2-D) projection images were produced by combining ray tracing techniques with automated image segmentation routines. The resultant images provide accurate and realistic volumetric representations of the Visible Human data set which is ultimately needed in medical virtual environment simulation. Ray tracing techniques provide methods by which 2-D volume views of a 3-D voxel array can be produced. The cross-sectional images can be scanned at different angles to produce rotated views of the voxel array. By combining volume views at incremental angles over 360 degrees a full volumetric representation of the voxel array, in this case the human male data set, can be computer generated and displayed without the speed and memory limitations of trying to display the entire data array. Additional texture and feature information can be obtained from the data by applying optical property equations to the ray scans. The imaging effects that can be added to volume renderings using these equations include shading, shadowing, and transparency. The automated segmentation routines provide a means to distinguish between various anatomical structures of the body. These routines can be used to differentiate between skin, fat, muscle, cartilage, blood vessels, and bone. By combining automated segmentation routines with the ray-tracing techniques, 2-D volume views of various anatomical structures and features can be isolated from the full data set. Examples of these segmentation abilities are demonstrated for the human male data set which include volume views of the skeletal systems, the musculoskeletal system, and part of the vascular system. The methods described above allow us to generate lifelike images, NURBS surface models, and realistic texture maps of specific anatomical structures. We have the capability to generate images that are both accurate and lifelike, much like photographic anatomical atlases. We can also generate images, models, and textures that have the clarity of medical artwork/illustrations, by highlighting the coloring of the ray traced structures with conventional colors instead of the natural color of the specimen. We are currently in the process of generating a comprehensive reference atlas of volume rendered images of the human body, soon to be published by Mosby-Year Book. The segmentation techniques needed to create this atlas also offer the accuracy and realism needed to create surface models and texture maps for a virtual environment for surgery simulation.

Anatomy, Cross-Sectional↗

PubMed1998

Automated image analysis system for detecting boundaries of live prostate cancer cells.

Image analysis provides a powerful tool for quantifying cell motility and has been used to correlate motility with metastatic potential in an animal model of prostate cancer. However, widespread use of this image analysis method has been limited because earlier methods of quantitative analysis required time-intensive and subjective manual tracing of cell contours. In this report, we describe a fully automated image segmentation algorithm for detection and morphometric description of prostatic cells. The segmentation system was tested on prostate cell images generated from Hoffman modulation contrast microscopy (47 cells at 64 time points = 3,008 images) and differential interference contrast microscopy (29 cells at 64 times points plus 1 cell at 62 time points = 1,918 images). Morphometric measurements were derived from computer-determined cell boundaries and compared with the same measurements derived from manually traced cell boundaries. Final correlation coefficients for area and perimeter measurements for Hoffman and differential interference contrast microscopy were (0.76, 0.62) and (0.93, 0.93), respectively. Results with our differential interference contrast images demonstrate that our segmentation algorithm reliably and efficiently replaces the need for manually traced cell boundaries in addition to eliminating intraobserver variation. Our automated segmentation process will have immediate utility in our motility analysis system that relates cell motility with metastatic potential of prostate cancer.

Adenocarcinoma↗

PubMed2003

Removal of reflections in the photographic assessment of PCO by fusion of digital retroillumination images.

PURPOSE: Automated image-analysis systems for objective assessment of posterior capsule opacification (PCO) depend on good image quality. One major drawback is the existence of light-reflection artifacts (Purkinje spots) in retroillumination images of the posterior capsule. Therefore, a software algorithm was developed that removes these artifacts by fusion of two or more digital images from the same eye, photographed in slightly different directions of gaze. METHODS: The image-fusion process comprises five steps: definition of a primary and a secondary image, automated segmentation of the region of interest and the Purkinje spots, manual selection of three pairs of corresponding points in both images, geometric registration and radiometric calibration of the regions to be inserted from the secondary image into the primary image. The program was tested with an image set of 30 eyes that had various degrees of PCO. A digital image acquisition system with a coaxial optical path was used to take retroillumination images from each eye in at least three different directions of gaze. RESULTS: In 28 cases all light-reflection artifacts within the capsulorrhexis rim could be removed entirely. In two cases, small parts of single Purkinje spots remained visible, because the reflections were located too closely in the primary and the secondary images. CONCLUSIONS: Fusion of digital retroillumination images provides high-quality, reflection-free PCO images. This allows objective PCO assessment systems to analyze 100% of the posterior capsule, leading to more accurate results.

Algorithms↗