Knowledge-based classification and tissue labeling of MR images of human brain

Brain tissue segmentation using improved kernelized rough-fuzzy C-means with spatio-contextual information from MRI

Segmentation of brain tissues from MRI often becomes crucial to properly investigate any region of the brain in order to detect abnormalities. However, the accurate segmentation of the brain tissues is a challenging task as the different tissue regions are usually imprecise, indiscernible, ambiguous, and overlapping. Additionally, different tissue regions are non-linearly separable. Noises and other artifacts may also present in the brain MRI. Therefore, conventional segmentation techniques may not often achieve desired accuracy. To deal those challenges, a robust kernelized rough fuzzy C-means clustering with spatial constraints (KRFCMSC) is proposed in this article for brain tissue segmentation. Here, the brain tissue segmentation from MRI is considered as a clustering of pixels problem. The basic idea behind the proposed technique is the judicious integration of the fuzzy set, rough set, and kernel trick along with spatial constraints (in the form of contextual information) to increase the clustering (segmentation) performance. The use of rough and fuzzy set theory in the clustering process handles the ambiguity, indiscernibility, vagueness and overlappingness of different brain tissue regions. While, the kernel trick increases the chance of linear separability of the complex regions which are otherwise not linearly separable in its original feature space. In order to deal the noisy pixels, here in the clustering process, the spatio-contextual information is introduced from the neighbouring pixels. Experiments are carried out on different real and synthetic benchmark brain MRI datasets (publicly available from Brainweb, and IBSR) without and with added noise. The performance of the proposed method is compared with five other counterpart clustering based segmentation techniques and evaluated using various supervised as well as unsupervised validity indices such as, overall accuracy, precision, recall, kappa, Jaccard, dice, and kernelized Xie-Beni index. Experimental results justify the superiority and robustness of the proposed method over other state-of-the-art methods on both benchmark real life and synthetic brain MRI datasets with and without added noise. Statistical significance of the better segmentation accuracy can be confirmed from the paired t-test results in favour of the proposed method compared to other counterpart methods.

Learning Based Segmentation of CT Brain Images: Application to Postoperative Hydrocephalic Scans

OBJECTIVE

Hydrocephalus is a medical condition in which there is an abnormal accumulation of cerebrospinal fluid (CSF) in the brain. Segmentation of brain imagery into brain tissue and CSF [before and after surgery, i.e., preoperative (pre-op) versus postoperative (post-op)] plays a crucial role in evaluating surgical treatment. Segmentation of pre-op images is often a relatively straightforward problem and has been well researched. However, segmenting post-op computational tomographic (CT) scans becomes more challenging due to distorted anatomy and subdural hematoma collections pressing on the brain. Most intensity- and feature-based segmentation methods fail to separate subdurals from brain and CSF as subdural geometry varies greatly across different patients and their intensity varies with time. We combat this problem by a learning approach that treats segmentation as supervised classification at the pixel level, i.e., a training set of CT scans with labeled pixel identities is employed.

METHODS

Our contributions include: 1) a dictionary learning framework that learns class (segment) specific dictionaries that can efficiently represent test samples from the same class while poorly represent corresponding samples from other classes; 2) quantification of associated computation and memory footprint; and 3) a customized training and test procedure for segmenting post-op hydrocephalic CT images.

RESULTS

Experiments performed on infant CT brain images acquired from the CURE Children's Hospital of Uganda reveal the success of our method against the state-of-the-art alternatives. We also demonstrate that the proposed algorithm is computationally less burdensome and exhibits a graceful degradation against a number of training samples, enhancing its deployment potential.

Improved Fuzzy C-Means based Particle Swarm Optimization (PSO) initialization and outlier rejection with level set methods for MR brain image segmentation

In this paper, a new image segmentation method based on Particle Swarm Optimization (PSO) and outlier rejection combined with level set is proposed. A traditional approach to the segmentation of Magnetic Resonance (MR) images is the Fuzzy C-Means (FCM) clustering algorithm. The membership function of this conventional algorithm is sensitive to the outlier and does not integrate the spatial information in the image. The algorithm is very sensitive to noise and in-homogeneities in the image, moreover, it depends on cluster centers initialization. To improve the outlier rejection and to reduce the noise sensitivity of conventional FCM clustering algorithm, a novel extended FCM algorithm for image segmentation is presented. In general, in the FCM algorithm the initial cluster centers are chosen randomly, with the help of PSO algorithm the clusters centers are chosen optimally. Our algorithm takes also into consideration the spatial neighborhood information. These a priori are used in the cost function to be optimized. For MR images, the resulting fuzzy clustering is used to set the initial level set contour. The results confirm the effectiveness of the proposed algorithm.

Fully automated segmentation of the pons and midbrain using human T1 MR brain images

Purpose

This paper describes a novel method to automatically segment the human brainstem into midbrain and pons, called LABS: Landmark-based Automated Brainstem Segmentation. LABS processes high-resolution structural magnetic resonance images (MRIs) according to a revised landmark-based approach integrated with a thresholding method, without manual interaction.

Methods

This method was first tested on morphological T1-weighted MRIs of 30 healthy subjects. Its reliability was further confirmed by including neurological patients (with Alzheimer's Disease) from the ADNI repository, in whom the presence of volumetric loss within the brainstem had been previously described. Segmentation accuracies were evaluated against expert-drawn manual delineation. To evaluate the quality of LABS segmentation we used volumetric, spatial overlap and distance-based metrics.

Results

The comparison between the quantitative measurements provided by LABS against manual segmentations revealed excellent results in healthy controls when considering either the midbrain (DICE measures higher that 0.9; Volume ratio around 1 and Hausdorff distance around 3) or the pons (DICE measures around 0.93; Volume ratio ranging 1.024–1.05 and Hausdorff distance around 2). Similar performances were detected for AD patients considering segmentation of the pons (DICE measures higher that 0.93; Volume ratio ranging from 0.97–0.98 and Hausdorff distance ranging 1.07–1.33), while LABS performed lower for the midbrain (DICE measures ranging 0.86–0.88; Volume ratio around 0.95 and Hausdorff distance ranging 1.71–2.15).

Conclusions

Our study represents the first attempt to validate a new fully automated method for in vivo segmentation of two anatomically complex brainstem subregions. We retain that our method might represent a useful tool for future applications in clinical practice.

Deformable templates guided discriminative models for robust 3D brain MRI segmentation

Automatically segmenting anatomical structures from 3D brain MRI images is an important task in neuroimaging. One major challenge is to design and learn effective image models accounting for the large variability in anatomy and data acquisition protocols. A deformable template is a type of generative model that attempts to explicitly match an input image with a template (atlas), and thus, they are robust against global intensity changes. On the other hand, discriminative models combine local image features to capture complex image patterns. In this paper, we propose a robust brain image segmentation algorithm that fuses together deformable templates and informative features. It takes advantage of the adaptation capability of the generative model and the classification power of the discriminative models. The proposed algorithm achieves both robustness and efficiency, and can be used to segment brain MRI images with large anatomical variations. We perform an extensive experimental study on four datasets of T1-weighted brain MRI data from different sources (1,082 MRI scans in total) and observe consistent improvement over the state-of-the-art systems.

Automatic MRI 2D brain segmentation using graph searching technique

Accurate and efficient segmentation of the whole brain in magnetic resonance (MR) images is a key task in many neuroscience and medical studies either because the whole brain is the final anatomical structure of interest or because the automatic extraction facilitates further analysis. The problem of segmenting brain MRI images has been extensively addressed by many researchers. Despite the relevant achievements obtained, automated segmentation of brain MRI imagery is still a challenging problem whose solution has to cope with critical aspects such as anatomical variability and pathological deformation. In the present paper, we describe and experimentally evaluate a method for segmenting brain from MRI images basing on two-dimensional graph searching principles for border detection. The segmentation of the whole brain over the entire volume is accomplished slice by slice, automatically detecting frames including eyes. The method is fully automatic and easily reproducible by computing the internal main parameters directly from the image data. The segmentation procedure is conceived as a tool of general applicability, although design requirements are especially commensurate with the accuracy required in clinical tasks such as surgical planning and post-surgical assessment. Several experiments were performed to assess the performance of the algorithm on a varied set of MRI images obtaining good results in terms of accuracy and stability.

A multidimensional segmentation evaluation for medical image data

Evaluation of segmentation methods is a crucial aspect in image processing, especially in the medical imaging field, where small differences between segmented regions in the anatomy can be of paramount importance. Usually, segmentation evaluation is based on a measure that depends on the number of segmented voxels inside and outside of some reference regions that are called gold standards. Although some other measures have been also used, in this work we propose a set of new similarity measures, based on different features, such as the location and intensity values of the misclassified voxels, and the connectivity and the boundaries of the segmented data. Using the multidimensional information provided by these measures, we propose a new evaluation method whose results are visualized applying a Principal Component Analysis of the data, obtaining a simplified graphical method to compare different segmentation results. We have carried out an intensive study using several classic segmentation methods applied to a set of MRI simulated data of the brain with several noise and RF inhomogeneity levels, and also to real data, showing that the new measures proposed here and the results that we have obtained from the multidimensional evaluation, improve the robustness of the evaluation and provides better understanding about the difference between segmentation methods.

Medical image segmentation using genetic algorithms

Genetic algorithms (GAs) have been found to be effective in the domain of medical image segmentation, since the problem can often be mapped to one of search in a complex and multimodal landscape. The challenges in medical image segmentation arise due to poor image contrast and artifacts that result in missing or diffuse organ/tissue boundaries. The resulting search space is therefore often noisy with a multitude of local optima. Not only does the genetic algorithmic framework prove to be effective in coming out of local optima, it also brings considerable flexibility into the segmentation procedure. In this paper, an attempt has been made to review the major applications of GAs to the domain of medical image segmentation.

Brain anatomical structure segmentation by hybrid discriminative/generative models

In this paper, a hybrid discriminative/generative model for brain anatomical structure segmentation is proposed. The learning aspect of the approach is emphasized. In the discriminative appearance models, various cues such as intensity and curvatures are combined to locally capture the complex appearances of different anatomical structures. A probabilistic boosting tree (PBT) framework is adopted to learn multiclass discriminative models that combine hundreds of features across different scales. On the generative model side, both global and local shape models are used to capture the shape information about each anatomical structure. The parameters to combine the discriminative appearance and generative shape models are also automatically learned. Thus, low-level and high-level information is learned and integrated in a hybrid model. Segmentations are obtained by minimizing an energy function associated with the proposed hybrid model. Finally, a grid-face structure is designed to explicitly represent the 3-D region topology. This representation handles an arbitrary number of regions and facilitates fast surface evolution. Our system was trained and tested on a set of 3-D magnetic resonance imaging (MRI) volumes and the results obtained are encouraging.

Automated segmentation and classification of high throughput yeast assay spots

Several technologies for characterizing genes and proteins from humans and other organisms use yeast growth or color development as read outs. The yeast two-hybrid assay, for example, detects protein-protein interactions by measuring the growth of yeast on a specific solid medium, or the ability of the yeast to change color when grown on a medium containing a chromogenic substrate. Current systems for analyzing the results of these types of assays rely on subjective and inefficient scoring of growth or color by human experts. Here, an image analysis system is described for scoring yeast growth and color development in high throughput biological assays. The goal is to locate the spots and score them in color images of two types of plates named "X-Gal" and "growth assay" plates, with uniformly placed spots (cell areas) on each plate (both plates in one image). The scoring system relies on color for the X-Gal spots, and texture properties for the growth assay spots. A maximum likelihood projection-based segmentation is developed to automatically locate spots of yeast on each plate. Then color histogram and wavelet texture features are extracted for scoring using an optimal linear transformation. Finally, an artificial neural network is used to score the X-Gal and growth assay spots using the extracted features. The performance of the system is evaluated using spots of 60 images. After training the networks using training and validation sets, the system was assessed on the test set. The overall accuracies of 95.4% and 88.2% are achieved, respectively, for scoring the X-Gal and growth assay spots.

MRI Fuzzy Segmentation of Brain Tissue Using Neighborhood Attraction With Neural-Network Optimization

Image segmentation is an indispensable process in the visualization of human tissues, particularly during clinical analysis of magnetic resonance (MR) images. Unfortunately, MR images always contain a significant amount of noise caused by operator performance, equipment, and the environment, which can lead to serious inaccuracies with segmentation. A robust segmentation technique based on an extension to the traditional fuzzy c-means (FCM) clustering algorithm is proposed in this paper. A neighborhood attraction, which is dependent on the relative location and features of neighboring pixels, is shown to improve the segmentation performance dramatically. The degree of attraction is optimized by a neural-network model. Simulated and real brain MR images with different noise levels are segmented to demonstrate the superiority of the proposed technique compared to other FCM-based methods. This segmentation method is a key component of an MR image-based classification system for brain tumors, currently being developed. Index Terms-Improved fuzzy c-means clustering (IFCM), magnetic resonance imaging (MRI), neighborhood attraction, segmentation.

Multiparametric ISODATA analysis of embolic stroke and rt-PA intervention in rat

To increase the sensitivity of MRI parameters to detect tissue damage of ischemic stroke, an unsupervised analysis method, Iterative Self-Organizing Data Analysis Technique Algorithm (ISODATA), was applied to analyze the temporal evolution of ischemic damage in a focal embolic cerebral ischemia model in rat with and without recombinant tissue plasminogen activator (rt-PA) treatment. Male Wistar rats subjected to embolic stroke were investigated using a 7-T MRI system. Rats were randomized into control (n=9) and treated (n=9) groups. The treated rats received rt-PA via a femoral vein at 4 h after onset of embolic ischemia. ISODATA analysis employed parametric maps or weighted images (T1, T2, and diffusion). ISODATA results with parametric maps are superior to ISODATA with weighted images, and both of them were highly correlated with the infarction size measured from the corresponding histological section. At 24 h after embolic stroke, the average map ISODATA lesion sizes were 37.7+/-7.0 and 39.2+/-5.6 mm2 for the treated and the control group, respectively. Average histological infarction areas were 37.9+/-7.4 mm2 for treated rats and 39.4+/-6.1 mm2 for controls. The R2 values of the linear correlation between map ISODATA and histological data were 0.98 and 0.96 for treated and control rats, respectively. Both histological and map ISODATA data suggest that there is no significant difference in infarction area between non-treated and rt-PA-treated rats when treatment was administered 4 h after the onset of embolic stroke. The ISODATA lesion size analysis was also sensitive to changes of lesion size during acute and subacute stages of stroke. Our data demonstrate that the multiparameter map ISODATA approach provides a more sensitive quantitation of the ischemic lesion at all time points than image ISODATA and single MRI parametric analysis using T1, T2 or ADCw.

Identifying multiple abdominal organs from CT image series using a multimodule contextual neural network and spatial fuzzy rules

Identifying abdominal organs is one of the essential steps in visualizing organ structure to assist in teaching, clinical training, diagnosis, and medical image retrieval. However, due to partial volume effects, gray-level similarities of adjacent organs, contrast media affect, and the relatively high variations of organ position and shape, automatically identifying abdominal organs has always been a high challenging task. To conquer these difficulties, this paper proposes combining a multimodule contextual neural network and spatial fuzzy rules and fuzzy descriptors for automatically identifying abdominal organs from a series of CT image slices. The multimodule contextual neural network segments each image slice through a divide-and-conquer concept, embedded within multiple neural network modules, where the results obtained from each module are forwarded to other modules for integration, in which contextual constraints are enforced. With this approach, the difficulties arising from partial volume effects, gray-level similarities of adjacent organs, and contrast media affect can be reduced to the extreme. To address the issue of high variations in organ position and shape, spatial fuzzy rules and fuzzy descriptors are adopted, along with a contour modification scheme implementing consecutive organ region overlap constraints. This approach has been tested on 40 sets of abdominal CT images, where each set consists of about 40 image slices. We have found that 99% of the organ regions in the test images are correctly identified as its belonging organs, implying the high promise of the proposed method.

Segmentation and visualization of nasopharyngeal carcinoma using MRI

In this study, a semi-automatic system was developed for nasopharyngeal carcinoma (NPC) tumor segmentation, volume measurement and visualization using magnetic resonance imaging (MRI). Some novel algorithms for tumor segmentation from MRI and inter-slice interpolation were integrated in this medical diagnosis system. This system was applied to 10 MR image data sets of NPC patients and satisfactory results were achieved. This system can be used as a clinical image analysis tool for doctors or radiologists to obtain tumor location from MRI, tumor volume estimation, and 3D information.

Automated cerebrum segmentation from three-dimensional sagittal brain MR images

We present a fully automated cerebrum segmentation algorithm for full three-dimensional sagittal brain MR images. First, cerebrum segmentation from a midsagittal brain MR image is performed utilizing landmarks, anatomical information, and a connectivity-based threshold segmentation algorithm as previously reported. Recognizing that cerebrum in laterally adjacent slices tends to have similar size and shape, we use the cerebrum segmentation result from the midsagittal brain MR image as a mask to guide cerebrum segmentation in adjacent lateral slices in an iterative fashion. This masking operation yields a masked image (preliminary cerebrum segmentation) for the next lateral slice, which may truncate brain region(s). Truncated regions are restored by first finding end points of their boundaries, by comparing the mask image and masked image boundaries, and then applying a connectivity-based algorithm. The resulting final extracted cerebrum image for this slice is then used as a mask for the next lateral slice. The algorithm yielded satisfactory fully automated cerebrum segmentations in three-dimensional sagittal brain MR images, and had performance superior to conventional edge detection algorithms for segmentation of cerebrum from 3D sagittal brain MR images.

Validation of partial tissue segmentation of single-channel magnetic resonance images of the brain

We describe and evaluate a practical, automated algorithm based on local statistical mixture modeling for segmenting single-channel, T1-weighted volumetric magnetic resonance images of the brain into gray matter, white matter, and cerebrospinal fluid. We employed a stereological sampling method to assess, prospectively, the performance of the method with respect to human experts on 10 normal T1-weighted brain scans acquired with a three-dimensional gradient echo pulse sequence. The overall kappa statistic for the concordance of the algorithm with the human experts was 0.806, while that among raters, excluding the algorithm, was 0.802. The algorithm had better agreement with the modal expert decision (kappa = 0.878). The algorithm could not be distinguished from the experts by this measure. We also validated the algorithm on a simulated MR scan of a digital brain phantom with known tissue composition. Global gray matter and white matter errors were 1% and <1%, respectively, and correlation coefficients with the underlying tissue model were 0.95 for gray matter, 0.98 for white matter, and 0.95 for cerebrospinal fluid. In both approaches to validation, we evaluated both local and global performance of the algorithm. Human experts generated slightly higher global gray matter proportion estimates on the test brain scans relative to the algorithm (3.7%) and on the simulated MR scan relative to the true tissue model (4.4%). The algorithm underestimated gray in some subcortical nuclei which contain admixed gray and white matter. We demonstrate the reliability of the method on individual 1 NEX data sets of the test subjects, and its insensitivity to the precise values of initial model parameters. The output of this algorithm is suitable for quantifying cerebral cortical tissue, using a commonly performed commercial pulse sequence.

A fuzzy logic approach to identifying brain structures in MRI using expert anatomic knowledge

We report a novel computer method for automatic labeling of structures in 3D MRI data sets using expert anatomical knowledge that is coded in fuzzy sets and fuzzy rules. The method first identifies major structures and then uses spatial relationships to these landmarks to recognize other structures. This labeling process simulates the iterative process that we ourselves use to locate structures in images. We demonstrate its application in three data sets, labeling brain MRI by locating the longitudinal and lateral fissures and the central sulci and then determining boundaries for the frontal lobes. Our method is adaptable to the identification of other anatomical structures.

Segmentation and interpretation of MR brain images: an improved active shape model

This paper reports a novel method for fully automated segmentation that is based on description of shape and its variation using point distribution models (PDM's). An improvement of the active shape procedure introduced by Cootes and Taylor to find new examples of previously learned shapes using PDM's is presented. The new method for segmentation and interpretation of deep neuroanatomic structures such as thalamus, putamen, ventricular system, etc. incorporates a priori knowledge about shapes of the neuroanatomic structures to provide their robust segmentation and labeling in magnetic resonance (MR) brain images. The method was trained in eight MR brain images and tested in 19 brain images by comparison to observer-defined independent standards. Neuroanatomic structures in all testing images were successfully identified. Computer-identified and observer-defined neuroanatomic structures agreed well. The average labeling error was 7%+/-3%. Border positioning errors were quite small, with the average border positioning error of 0.8+/-0.1 pixels in 256 x 256 MR images. The presented method was specifically developed for segmentation of neuroanatomic structures in MR brain images. However, it is generally applicable to virtually any task involving deformable shape analysis.

Understanding coronary artery movement: a knowledge-based approach

The aim of this paper is to describe a knowledge-based system that interprets three-dimensional (3D) coronary artery movement, using data from digital subtraction angiography image sequences. Dynamic information obtained from artery centerline 3D reconstruction and optical flow estimation, is classified according to experimental evidence indicating that artery displacements are quasi-homogeneous by a segment analysis. Characteristic motion features like displacement direction, perpendicular/radial components, rotation direction, curvature and torsion are qualitatively described from an image sequence using symbolic labels. These facts are then related and interpreted using anatomical-functional knowledge provided by a specialist, as well as spatial and temporal knowledge, applying spatio-temporal reasoning schemes. Facts, knowledge and reasoning rules are stated in a declarative form. Detailed examples of local and global interpretation results, using a real reconstructed angiographic biplane image sequence are presented in order to illustrate how our system suitably interprets coronary artery dynamic behavior.

Automatic tumor segmentation using knowledge-based techniques

A system that automatically segments and labels glioblastoma-multiforme tumors in magnetic resonance images (MRI's) of the human brain is presented. The MRI's consist of T1-weighted, proton density, and T2-weighted feature images and are processed by a system which integrates knowledge-based (KB) techniques with multispectral analysis. Initial segmentation is performed by an unsupervised clustering algorithm. The segmented image, along with cluster centers for each class are provided to a rule-based expert system which extracts the intracranial region. Multispectral histogram analysis separates suspected tumor from the rest of the intracranial region, with region analysis used in performing the final tumor labeling. This system has been trained on three volume data sets and tested on thirteen unseen volume data sets acquired from a single MRI system. The KB tumor segmentation was compared with supervised, radiologist-labeled "ground truth" tumor volumes and supervised k-nearest neighbors tumor segmentations. The results of this system generally correspond well to ground truth, both on a per slice basis and more importantly in tracking total tumor volume during treatment over time.

MRI measurement of brain tumor response: comparison of visual metric and automatic segmentation

An automatic magnetic resonance imaging (MRI) multispectral segmentation method and a visual metric are compared for their effectiveness to measure tumor response to therapy. Automatic response measurements are important for multicenter clinical trials. A visual metric such as the product of the largest diameter and the largest perpendicular diameter of the tumor is a standard approach, and is currently used in the Radiation Treatment Oncology Group (RTOG) and the Eastern Cooperative Oncology Group (EGOG) clinical trials. In the standard approach, the tumor response is based on the percentage change in the visual metric and is categorized into cure, partial response, stable disease, or progression. Both visual and automatic methods are applied to six brain tumor cases (gliomas) of varying levels of segmentation difficulty. The analyzed data were serial multispectral MR images, collected using MR contrast enhancement. A fully automatic knowledge guided method (KG) was applied to the MRI multispectral data, while the visual metric was taken from the MRI films using the T1 gadolinium enhanced image, with repeat measurements done by two radiologists and two residents. Tumor measurements from both visual and automatic methods are compared to "ground truth," (GT) i.e., manually segmented tumor. The KG method was found to slightly overestimate tumor volume, but in a consistent manner, and the estimated tumor response compared very well to hand-drawn ground truth with a correlation coefficient of 0.96. In contrast, the visually estimated metric had a large variation between observers, particularly for difficult cases, where the tumor margins are not well delineated. The inter-observer variation for the measurement of the visual metric was only 16%, i.e., observers generally agreed on the lengths of the diameters. However, in 30% of the studied cases no consensus was found for the categorical tumor response measurement, indicating that the categories are very sensitive to variations in the diameter measurements. Moreover, the method failed to correctly identify the response in half of the cases. The data demonstrate that automatic 3D methods are clearly necessary for objective and clinically meaningful assessment of tumor volume in single or multicenter clinical trials.

Fully automatic segmentation of the brain in MRI

A robust fully automatic method for segmenting the brain from head magnetic resonance (MR) images has been developed, which works even in the presence of radio frequency (RF) inhomogeneities. It has been successful in segmenting the brain in every slice from head images acquired from several different MRI scanners, using different-resolution images and different echo sequences. The method uses an integrated approach which employs image processing techniques based on anisotropic filters and "snakes" contouring techniques, and a priori knowledge, which is used to remove the eyes, which are tricky to remove based on image intensity alone. It is a multistage process, involving first removal of the background noise leaving a head mask, then finding a rough outline of the brain, then refinement of the rough brain outline to a final mask. The paper describes the main features of the method, and gives results for some brain studies.

Medical image analysis with fuzzy models

This paper updates several recent surveys on the use of fuzzy models for segmentation and edge detection in medical image data. Our survey is divided into methods based on supervised and unsupervised learning (that is, on whether there are or are not labelled data available for supervising the computations), and is organized first and foremost by groups (that we know of!) that are active in this area. Our review is aimed more towards 'who is doing it' rather than 'how good it is'. This is partially dictated by the fact that direct comparisons of supervised and unsupervised methods is somewhat akin to comparing apples and oranges. There is a further subdivision into methods for two- and three-dimensional data and/or problems. We do not cover methods based on neural-like networks or fuzzy reasoning systems. These topics are covered in a recently published companion survey by keller et al.

Monitoring brain tumor response to therapy using MRI segmentation

The performance evaluation of a semi-supervised fuzzy c-means (SFCM) clustering method for monitoring brain tumor volume changes during the course of routine clinical radiation-therapeutic and chemo-therapeutic regimens is presented. The tumor volume determined using the SFCM method was compared with the volume estimates obtained using three other methods: (a) a k nearest neighbor (kNN) classifier, b) a grey level thresholding and seed growing (ISG-SG) method and c) a manual pixel labeling (GT) method for ground truth estimation. The SFCM and kNN methods are applied to the multispectral, contrast enhanced T1, proton density, and T2 weighted, magnetic resonance images (MRI) whereas the ISG-SG and GT methods are applied only to the contrast enhanced T1 weighted image. Estimations of tumor volume were made on eight patient cases with follow-up MRI scans performed over a 32 week interval during treatment. The tumor cases studied include one meningioma, two brain metastases and five gliomas. Comparisons with manually labeled ground truth estimations showed that there is a limited agreement between the segmentation methods for absolute tumor volume measurements when using images of patients after treatment. The average intraobserver reproducibility for the SFCM, kNN and ISG-SG methods was found to be 5.8%, 6.6% and 8.9%, respectively. The average of the interobserver reproducibility of these methods was found to be 5.5%, 6.5% and 11.4%, respectively. For the measurement of relative change of tumor volume as required for the response assessment, the multi-spectral methods kNN and SFCM are therefore preferred over the seedgrowing method.

Normal brain volume measurements using multispectral MRI segmentation

The performance of a supervised k-nearest neighbor (kNN) classifier and a semisupervised fuzzy c-means (SFCM) clustering segmentation method are evaluated for reproducible measurement of the volumes of normal brain tissues and cerebrospinal fluid. The stability of the two segmentation methods is evaluated for (a) operator selection of training data, (b) reproducibility during repeat imaging sessions to determine any variations in the sensor performance over time, (c) variations in the measured volumes between different subjects, and (d) variability with different imaging parameters. The variations were found to be dependent on the type of measured tissue and the operator performing the segmentations. The variability during repeat imaging sessions for the SFCM method was < 3%. The absolute volumes of the brain matter and cerebrospinal fluid between subjects varied quite large, ranging from 9% to 13%. The intraobserver and interobserver reproducibility for SFCM were < 4% for the soft tissues and 6% for cerebrospinal fluid. The corresponding results for the kNN segmentation method were higher compared to the SFCM method.

Knowledge-based interpretation of MR brain images

The authors have developed a method for fully automated segmentation and labeling of 17 neuroanatomic structures such as thalamus, caudate nucleus, ventricular system, etc. in magnetic resonance (MR) brain images. The authors' method is based on a hypothesize-and-verify principle and uses a genetic algorithm (GA) optimization technique to generate and evaluate image interpretation hypotheses in a feedback loop. The authors' method was trained in 20 individual T1-weighted MR images. Observer-defined contours of neuroanatomic structures were used as a priori knowledge. The method's performance was validated in eight MR images by comparison to observer-defined independent standards. The GA-based image interpretation method correctly interpreted neuroanatomic structures in all images from the test set. Computer-identified and observer-defined neuroanatomic structure areas correlated very well (r=0.99, y=0,95x-2.1). Border positioning errors were small, with a root mean square (rms) border positioning error of 1.5+/-0.6 pixels. The authors' GA-based image interpretation method represents a novel approach to image interpretation and has been shown to produce accurate labeling of neuroanatomic structures in a set of MR brain images.

Labeling of MR brain images using Boolean neural network

Presents a knowledge-based approach for labeling two-dimensional (2-D) magnetic resonance (MR) brain images using the Boolean neural network (BNN), which has binary inputs and outputs, integer weights, fast learning and classification, and guaranteed convergence. The approach consists of two components: a BNN clustering algorithm and a constraint satisfying Boolean neural network (CSBNN) labeling procedure. The BNN clustering algorithm is developed to initially segment an image into a number of regions. Then the segmented regions are labeled with the CSBNN, which is a modified version of BNN. The CSBNN uses a knowledge base that contains information on image-feature space and tissue models as constraints. The method is tested using sets of MR brain images. The regions of the different brain tissues are satisfactorily segmented and labeled. A comparison with the Hopfield neural network and the traditional simulated annealing method for image labeling is provided. The comparison results show that the CSBNN approach offers a fast, feasible, and reliable alternative to the existing techniques for medical image labeling.

MRI segmentation: methods and applications

The current literature on MRI segmentation methods is reviewed. Particular emphasis is placed on the relative merits of single image versus multispectral segmentation, and supervised versus unsupervised segmentation methods. Image pre-processing and registration are discussed, as well as methods of validation. The application of MRI segmentation for tumor volume measurements during the course of therapy is presented here as an example, illustrating problems associated with inter- and intra-observer variations inherent to supervised methods.

Comparison of supervised MRI segmentation methods for tumor volume determination during therapy

Two different multispectral pattern recognition methods are used to segment magnetic resonance images (MRI) of the brain for quantitative estimation of tumor volume and volume changes with therapy. A supervised k-nearest neighbor (kNN) rule and a semi-supervised fuzzy c-means (SFCM) method are used to segment MRI slice data. Tumor volumes as determined by the kNN and SFCM segmentation methods are compared with two reference methods, based on image grey scale, as a basis for an estimation of ground truth, namely: (a) a commonly used seed growing method that is applied to the contrast enhanced T1-weighted image, and (b) a manual segmentation method using a custom-designed graphical user interface applied to the same raw image (T1-weighted) dataset. Emphasis is placed on measurement of intra and inter observer reproducibility using the proposed methods. Intra- and interobserver variation for the kNN method was 9% and 5%, respectively. The results for the SFCM method was a little better at 6% and 4%, respectively. For the seed growing method, the intra-observer variation was 6% and the interobserver variation was 17%, significantly larger when compared with the multispectral methods. The absolute tumor volume determined by the multispectral segmentation methods was consistently smaller than that observed for the reference methods. The results of this study are found to be very patient case-dependent. The results for SFCM suggest that it should be useful for relative measurements of tumor volume during therapy, but further studies are required. This work demonstrates the need for minimally supervised or unsupervised methods for tumor volume measurements.

Object recognition in brain CT-scans: knowledge-based fusion of data from multiple feature extractors

Describes a knowledge-based image interpretation system for the segmentation and labeling of a series of 2-D brain X-ray CT-scans, parallel to the orbito-meatal plane. The system combines the image primitive information produced by different low level vision techniques in order to improve the reliability of the segmentation and the image interpretation. It is implemented in a blackboard environment that is holding various types of prior information and which controls the interpretation process. The scoring model is applied for the fusion of information derived from three types of image primitives (points, edges, and regions). A model, containing both analogical and propositional knowledge on the brain objects, is used to direct the interpretation process. The linguistic variables, introduced to describe the propositional features of the brain model, are defined by fuzzy membership functions. Constraint functions are applied to evaluate the plausibility of the mapping between image primitives and brain model data objects. Procedural knowledge has been integrated into different knowledge sources. Experimental results illustrate the reliability and robustness of the system against small variations in slice orientation and interpatient variability in the images.

Segmentation of brain parenchyma and cerebrospinal fluid in multispectral magnetic resonance images

Presents a new method to segment brain parenchyma and cerebrospinal fluid spaces automatically in routine axial spin echo multispectral MR images. The algorithm simultaneously incorporates information about anatomical boundaries (shape) and tissue signature (grey scale) using a priori knowledge. The head and brain are divided into four regions and seven different tissue types. Each tissue type c is modeled by a multivariate Gaussian distribution N(mu(c),Sigma(c)). Each region is associated with a finite mixture density corresponding to its constituent tissue types. Initial estimates of tissue parameters {mu(c),Sigma(c )}(c=1,...,7) are obtained from k-means clustering of a single slice used for training. The first algorithmic step uses the EM-algorithm for adjusting the initial tissue parameter estimates to the MR data of new patients. The second step uses a recently developed model of dynamic contours to detect three simply closed nonintersecting curves in the plane, constituting the arachnoid/dura mater boundary of the brain, the border between the subarachnoid space and brain parenchyma, and the inner border of the parenchyma toward the lateral ventricles. The model, which is formulated by energy functions in a Bayesian framework, incorporates a priori knowledge, smoothness constraints, and updated tissue type parameters. Satisfactory maximum a posteriori probability estimates of the closed contour curves defined by the model were found using simulated annealing.