Search

Accurately measuring nonlinear effective connectivity is a crucial step in investigating brain functions. Brain signals like EEG is nonstationary. Many effective connectivity methods have been proposed but they have drawbacks in their models such as a weakness in proposing a way for hyperparameter and time lag selection as well as dealing with non-stationarity of the time series. This paper proposes an effective connectivity model based on a hybrid neural network model which uses Empirical Wavelet Transform (EWT) and a long short-term memory network (LSTM). The best hyperparameters and time lag are selected using Bayesian Optimization (BO). Due to the importance of generalizability in neural networks and calculating GC, an algorithm was proposed to choose the best generalizable weights. The model was evaluated using simulated and real EEG data consisting of attention deficit hyperactivity disorder (ADHD) and healthy subjects. The proposed model's performance on simulated data was evaluated by comparing it with other neural networks, including LSTM, CNN-LSTM, GRU, RNN, and MLP, using a Blocked cross-validation approach. GC of the simulated data was compared with GRU, linear Granger causality (LGC), Kernel Granger Causality (KGC), Partial Directed Coherence (PDC), and Directed Transfer Function (DTF). Our results demonstrated that the proposed model was superior to the mentioned models. Another advantage of our model is robustness against noise. The results showed that the proposed model can identify the connections in noisy conditions. The comparison of the effective connectivity of ADHD and the healthy group showed that the results are in accordance with previous studies.

To estimate dynamic functional connectivity (dFC), the conventional method of sliding window correlation (SWC) suffers from poor performance of dynamic connection detection. This stems from the equal weighting of observations, suboptimal time scale, nonsparse output, and the fact that it is bivariate. To overcome these limitations, we exploited the kernel-reweighted logistic regression (KELLER) algorithm, a method that is common in genetic studies, to estimate dFC in resting state functional magnetic resonance imaging (rs-fMRI) data. KELLER can estimate dFC through estimating both spatial and temporal patterns of functional connectivity between brain regions. This paper compares the performance of the proposed KELLER method with current methods (SWC and tapered-SWC (T-SWC) with different window lengths) based on both simulated and real rs-fMRI data. Estimated dFC networks were assessed for detecting dynamically connected brain region pairs with hypothesis testing. Simulation results revealed that KELLER can detect dynamic connections with a statistical power of 87.35% compared with 70.17% and 58.54% associated with T-SWC (p-value = .001) and SWC (p-value <.001), respectively. Results of these different methods applied on real rs-fMRI data were investigated for two aspects: calculating the similarity between identified mean dynamic pattern and identifying dynamic pattern in default mode network (DMN). In 68% of subjects, the results of T-SWC with window length of 100 s, among different window lengths, demonstrated the highest similarity to those of KELLER. With regards to DMN, KELLER estimated previously reported dynamic connection pairs between dorsal and ventral DMN while SWC-based method was unable to detect these dynamic connections., (© 2020 The Authors. Human Brain Mapping published by Wiley Periodicals LLC.)

It has been found that specific regions in the brain are dedicated to specific functions. Detection and analysis of the constituent functional networks of the brain is of great importance for understanding the brain functionality and diagnosing some neuropsychiatric illnesses. In this paper, we introduce Non-negative Tensor Factorization (NTF) methods to identify the overlapping communities in brain networks using resting-state functional Magnetic Resonance Imaging (rs-fMRI) data. Instead of taking average over a group of subjects, we use individual subject connectivity matrices to build the tensor data. Decomposed factors indicate the community membership probabilities and inter-subject variability indices modeling the community strengths over subjects. In contrast to the methods based on Non-negative Matrix Factorization (NMF) which are generally applied to the average connectivity matrices, using tensor factorization modeling preserves the information conveyed by the individual subjects. The experiments are carried out on simulated data as well as real Human Connectome Project (HCP) rs-fMRI datasets. To evaluate the effectiveness of the proposed framework, we have computed reproducibility over time and groups of subjects. Test-retest reliability is also examined through computing the intra-class correlation coefficient (ICC) index. The results show that the proposed NTF-based frameworks lead to stable and accurate results., (Copyright © 2019 Elsevier B.V. All rights reserved.)

Prediction of sudden cardiac death continues to gain universal attention as a promising approach to saving millions of lives threatened by sudden cardiac death (SCD). This study attempts to promote the literature from mere feature extraction analysis to developing strategies for manipulating the extracted features to target improvement of classification accuracy. To this end, a novel approach to local feature subset selection is applied using meticulous methodologies developed in previous studies of this team for extracting features from non-linear, time-frequency, and classical processes. We are therefore enabled to select features that differ from one another in each 1-min interval before the incident. Using the proposed algorithm, SCD can be predicted 12 min before the onset; thus, more propitious results are achieved. Additionally, through defining a utility function and employing statistical analysis, the alarm threshold has effectively been determined as 83%. Having selected the best combination of features, the two classes are classified using the multilayer perceptron (MLP) classifier. The most effective features would subsequently be discussed considering their prevalence in the rank-based selection. The results indicate the significant capacity of the proposed method for predicting SCD as well as selecting the appropriate processing method at any time before the incident. Graphical abstract ᅟ.

Wireless capsule endoscopy (WCE) is a great breakthrough for Gastrointestinal (GI) Tract diagnoses which can view the entire gastrointestinal tract, especially the small bowel, without invasiveness and sedation. However, a tough problem associated with this new device is that too many images to be inspected by naked eyes is difficult for physicians, Thus it is essential to find an automatic and intelligent diagnosis method to help physicians. In this paper, a new segmentation algorithm for detection of Lymphangiectasia, Xanthoma, Crohn, and Stenosis in WCE images is proposed. This new approach mainly uses the HSV color space, sigmoid function and canny edge detector. We compare our method with a fuzzy c-mean clustering. We show that sensitivities of the sigmoid function for Lymphangiectasia, Lymphoid hyperplasia, severe Crohn's disease, Xanthoma and ulcerated Stenosis are respectively 89.32%, 91.27%, 95.45%, 87.01%, 97% and sensitivities of the fuzzy c-means clustering with same order are 83.91%, 86.7%, 96.38%, 90.4%, 93.83%. Totally, the sigmoid function is more specific and sensitive, with same accuracy.

Diffusion tensor magnetic resonance imaging (DTMRI) is a non-invasive tool for the investigation of white matter structure within the brain. However, the traditional tensor model is unable to characterize anisotropies of orders higher than two in heterogeneous areas containing more than one fiber population. To resolve this issue, high angular resolution diffusion imaging (HARDI) with a large number of diffusion encoding gradients is used along with reconstruction methods such as Q-ball. Using HARDI data, the fiber orientation distribution function (ODF) on the unit sphere is calculated and used to extract the principal diffusion directions (PDDs). Fast and accurate estimation of PDDs is a prerequisite for tracking algorithms that deal with fiber crossings. In this paper, the PDDs are defined as the directions around which the ODF data is concentrated. Estimates of the PDDs based on this definition are less sensitive to noise in comparison with the previous approaches. A clustering approach to estimate the PDDs is proposed which is an extension of fuzzy c-means clustering developed for orientation of points on a sphere. MDL (Minimum description length) principle is proposed to estimate the number of PDDs. Using both simulated and real diffusion data, the proposed method has been evaluated and compared with some previous protocols. Experimental results show that the proposed clustering algorithm is more accurate, more resistant to noise, and faster than some of techniques currently being utilized., (Copyright © 2011 Elsevier Inc. All rights reserved.)

This paper proposes image processing algorithms to recognize five types of white blood cells in peripheral blood automatically. First, a method based on Gram-Schmidt orthogonalization is proposed along with a snake algorithm to segment nucleus and cytoplasm of the cells. Then, a variety of features are extracted from the segmented regions. Next, most discriminative features are selected using a Sequential Forward Selection (SFS) algorithm and performances of two classifiers, Artificial Neural Network (ANN) and Support Vector Machine (SVM), are compared. The results demonstrate that the proposed methods are accurate and sufficiently fast to be used in hematological laboratories., (Copyright © 2011 Elsevier Ltd. All rights reserved.)

We propose a novel approach for evaluating the performance of activation detection in real (experimental) datasets using a new mutual information (MI)-based metric and compare its sensitivity to several existing performance metrics in both simulated and real datasets. The proposed approach is based on measuring the approximate MI between the fMRI time-series of a validation dataset and a calculated activation map (thresholded label map or continuous map) from an independent training dataset. The MI metric is used to measure the amount of information preserved during the extraction of an activation map from experimentally related fMRI time-series. The processing method that preserves maximal information between the maps and related time-series is proposed to be superior. The results on simulation datasets for multiple analysis models are consistent with the results of ROC curves, but are shown to have lower information content than for real datasets, limiting their generalizability. In real datasets for group analyses using the general linear model (GLM; FSL4 and SPM5), we show that MI values are (1) larger for groups of 15 versus 10 subjects and (2) more sensitive measures than reproducibility (for continuous maps) or Jaccard overlap metrics (for thresholded maps). We also show that (1) for an increasing fraction of nominally active voxels, both MI and false discovery rate (FDR) increase, and (2) at a fixed FDR, GLM using FSL4 tends to extract more voxels and more information than SPM5 using the default processing techniques in each package., (Copyright © 2010 Wiley-Liss, Inc.)

Purpose: To develop an automatic atlas-based method for segmentation of fiber bundles using High Angular Resolution Diffusion Imaging (HARDI) data., Hypothesis: Quantitative evaluation of diffusion characteristics inside specific fiber bundles provides new insights into disease developments, evolutions, therapy effects, and surgical interventions., Background: Most of previous segmentation methods use similarity measures and strategies that do not lead to accurate segmentation results. They also suffer from subjectivity of initial seeds and regions of interest (ROI) defined by operator., Materials and Methods: We propose a novel method that uses Spherical Harmonic Coefficients (SHC) of HARDI diffusion signals to compute Orientation Distribution Function (ODF) and to extract Principal Diffusion Directions (PDDs). The proposed method selects most collinear PDD of neighbors of each voxel. Then, based on SHC and selected PDD, a similarity measure is proposed and used as a speed function in the level set framework that segments fiber bundles. To automate the process, an atlas-based method is used to select initial seeds for fiber bundles. To generate data for evaluation of the proposed method, an artificial pattern consisting of three crossing bundles intersected by a circular bundle is created. Also, two normal controls are imaged by two different HARDI protocols., Results: Segmentation results for different fiber bundles in simulated data and normal control data are presented. Influence of threshold selection on the proposed segmentation method is evaluated using Dice coefficient. Also, effect of increasing the number of gradient directions on accuracy of extracted PDDs is evaluated., Conclusion: The proposed segmentation method has advantages over previous methods especially those that use similarity measures based on diffusion tensor imaging (DTI) data. These advantages are achieved by proper propagation of a hyper-surface in fiber crossing areas without making assumptions about diffusivity profile and selection of initial seeds or ROI., (Copyright © 2010 Elsevier Inc. All rights reserved.)

The hippocampus has been the primary region of interest in the preoperative imaging investigations of mesial temporal lobe epilepsy (mTLE). Hippocampal imaging and electroencephalographic features may be sufficient in several cases to declare the epileptogenic focus. In particular, hippocampal atrophy, as appreciated on T1-weighted (T1W) magnetic resonance (MR) images, may suggest a mesial temporal sclerosis. Qualitative visual assessment of hippocampal volume, however, is influenced by head position in the magnet and the amount of atrophy in different parts of the hippocampus. An entropy-based segmentation algorithm for subcortical brain structures (LocalInfo) was developed and supplemented by both a new multiple atlas strategy and a free-form deformation step to capture structural variability. Manually segmented T1-weighted magnetic resonance (MR) images of 10 non-epileptic subjects were used as atlases for the proposed automatic segmentation protocol which was applied to a cohort of 46 mTLE patients. The segmentation and lateralization accuracies of the proposed technique were compared with those of two other available programs, HAMMER and FreeSurfer, in addition to the manual method. The Dice coefficient for the proposed method was 11% (p<10(-5)) and 14% (p<10(-4)) higher in comparison with the HAMMER and FreeSurfer, respectively. Mean and Hausdorff distances in the proposed method were also 14% (p<0.2) and 26% (p<10(-3)) lower in comparison with HAMMER and 8% (p<0.8) and 48% (p<10(-5)) lower in comparison with FreeSurfer, respectively. LocalInfo proved to have higher concordance (87%) with the manual segmentation method than either HAMMER (85%) or FreeSurfer (83%). The accuracy of lateralization by volumetry in this study with LocalInfo was 74% compared to 78% with the manual segmentation method. LocalInfo yields a closer approximation to that of manual segmentation and may therefore prove to be more reliable than currently published automatic segmentation algorithms., (Copyright © 2010 Elsevier Inc. All rights reserved.)

Purpose: The authors propose a fast, robust, nonparametric, entropy-based, coupled, multishape approach to segment subcortical brain structures from magnetic resonance images (MRIs)., Methods: The proposed method uses three types of information: Image intensity, tissue types, and locations of structures. The image intensity information is captured by estimating the probability density function (pdf) of the image intensities in each structure. The tissue type information is captured by applying an unsupervised tissue segmentation method to the image and estimating a probability mass function (pmf) for the tissue type of each structure. The location information is captured by estimating pdf of the location of each structure from the training datasets. The resulting pmf's and pdf's are used to define an entropy function whose minimum corresponds to a desirable segmentation of the structures. The authors propose a three-step optimization strategy for the segmentation method. In the first step, a powerful automatic initialization method is developed based on tissue type and location information of the structures. In the second step, a quasi-Newton method is used to optimize the parameters of the energy function. To speed up the iterations, derivatives of the energy function with respect to its parameters are analytically derived and used in the optimization process. In the last step, the limitations related to the prior shape model are removed and a level-set method is applied for the fine tuning of the segmentation results., Results: The proposed method is applied to two different datasets and the results are compared to those of previous methods in literature. Experimental results are presented for lateral ventricles, caudate, thalamus, putamen, pallidum, hippocampus, and amygdala., Conclusions: The results illustrate superior performance of the proposed segmentation method compared to other methods in literature. The execution time of the algorithm is a few minutes, suitable for a variety of applications.

Super-paramagnetic iron oxide (SPIO) nanoparticles are actively investigated to enhance disease detection through molecular imaging using magnetic resonance imaging (MRI). Detection of the cells labeled by SPIO depends on the MRI protocols and pulse sequence parameters that can be optimized. To evaluate the sensitivity and specificity of the image acquisition methods and to obtain optimal imaging parameters for single-cell detection, we further developed an MRI simulator. The simulator models an object (tissue) at a microscopic level to evaluate effects of spatial distribution and concentration of nanoparticles on the resulting image. In this study, the simulator was used to evaluate and compare imaging of the labeled cells by the gradient-echo (GE), true-FISP [fast imaging employing steady-state acquisition (FIESTA)] and echo-planar imaging (EPI) pulse sequences. Effects of the imaging and object parameters, such as field strength, imaging protocol and pulse sequence parameters, imaging resolution, cell iron load, position of SPIO within the voxel and cell division within the voxel, were investigated in the work. The results suggest that true-FISP has the highest sensitivity for single-cell detection by MRI., (Copyright 2010 Elsevier Inc. All rights reserved.)

While information about anatomy is available in CT images, information about physiology and metabolism is available in PET images. To integrate both information, the two images are fused. Image fusion methods include simple methods like pixel averaging and sophisticated methods like wavelet transformation. An advantage of using wavelet transformation is that it preserves significant parts of each image. After creating lesions of 10, 8, 6 mm in a NURBS (non-uniform rational B-splines) based cardiac torso (NCAT) phantom, PET images were simulated using SimSET simulator. Attenuation maps of the activity phantom were used as CT images. Each of the PET and CT images was divided into an approximation image and three detailed images by the wavelet transform. The corresponding transformed images generated from the PET and CT images were fused in nine different ways to generate composite images, which were compared to the original images. The basis of comparison is the lesion-to-tissue contrast in the fused image in comparison to the lesion-to-tissue contrast in the original PET and CT images. Our results showed that except for one method, the lesion-to-tissue contrast in the fused image was higher than that of the CT images. In the first six methods, the lesion-to-tissue contrast in the fused image was less than the contrast, in the PET image. In the other three methods, the contrast in the fused image was higher than in the PET image. This was true in cases of 10, 8, 6 mm lesions. In conclusion, we have show that the approximation image produced a better ultimate image and that the lesion-to-tissue contrast in the fused image was also better than that of the original PET and CT images. This is because the approximation image is comprised of fundamental information of the signal (low frequency) that directly affects the image contrast.

We propose an automatic method for the segmentation of the brain structures in three dimensional (3D) Magnetic Resonance Images (MRI). The proposed method consists of two stages. In the first stage, we represent the shape of the structure using Geometric Moment Invariants (GMIs) in 8 scales. For each scale, an Artificial Neural Network (ANN) is designed to approximate the signed distance function of a desired structure. The GMIs along with the voxel intensities and coordinates are used as the input features of the ANN and the signed distance function as its output. In the second stage, we combine the outputs of the ANNs of the first stage and design another ANN to classify the image voxels into two classes, inside or outside of the structure. We introduce a fast method for moment calculations. The proposed method is applied to the segmentation of caudate, putamen, and thalamus in MRI where it has outperformed other methods in the literature.

The main objective of this paper is to present methods and results for the estimation of parameters of our proposed integrated magnetoencephalography (MEG) and functional magnetic resonance imaging (fMRI) model. We use real auditory MEG and fMRI datasets from 7 normal subjects to estimate the parameters of the model. The MEG and fMRI data were acquired at different times, but the stimulus profile was the same for both techniques. We use independent component analysis (ICA) to extract activation-related signal from the MEG data. The stimulus-correlated ICA component is used to estimate MEG parameters of the model. The temporal and spatial information of the fMRI datasets are used to estimate fMRI parameters of the model. The estimated parameters have reasonable means and standard deviations for all subjects. Goodness of fit of the real data to our model shows the possibility of using the proposed model to simulate realistic datasets for evaluation of integrated MEG/fMRI analysis methods.

Background: The purpose of this work was to evaluate effects of Gd-diethylenetriaminepentacetic acid (DTPA) injection on T(1)-weighted images of stroke and lesion segmentation and characterization results generated by our multiparametric iterative self-organizing data (ISODATA) method. The post-Gd image incorporates vasculature information into the analysis., Methods: Either a pre-Gd T(1)-weighted image (T1WI) or a post-Gd T1WI was used along with diffusion-, T(2)- and proton-density-weighted images in the analysis. ISODATA is a data-driven method that segments and characterizes tissue damage in stroke using multiparametric MRI., Results: Experimental results in both animal and human studies showed that the use of post-Gd T1WI modified the segmentation and characterization results on the periphery of the lesion. The peripheral region that changes with Gd-DTPA has a higher permeability compared to the rest of the lesion. Either of the data sets (including pre- or post-Gd T1WI) was used to estimate the tissue recovery and generated consistent results., Conclusions: This study shows that our multiparametric ISODATA approach consistently identifies and characterizes the core of the ischemic lesion. It also shows that the inclusion of post-Gd T1WI results in the segmentation and characterization of the lesion periphery if it has a higher permeability compared to the rest of the lesion. Finally, it confirms that the multiparametric ISODATA MRI characterizes tissue damage and recovery in stroke.

A new rotation-invariant texture-analysis technique using Radon and wavelet transforms is proposed. This technique utilizes the Radon transform to convert the rotation to translation and then applies a translation-invariant wavelet transform to the result to extract texture features. A kappa-nearest neighbors classifier is employed to classify texture patterns. A method to find the optimal number of projections for the Radon transform is proposed. It is shown that the extracted features generate an efficient orthogonal feature space. It is also shown that the proposed features extract both of the local and directional information of the texture patterns. The proposed method is robust to additive white noise as a result of summing pixel values to generate projections in the Radon transform step. To test and evaluate the method, we employed several sets of textures along with different wavelet bases. Experimental results show the superiority of the proposed method and its robustness to additive white noise in comparison with some recent texture-analysis methods.

This paper presents a new approach to rotation invariant texture classification. The proposed approach benefits from the fact that most of the texture patterns either have directionality (anisotropic textures) or are not with a specific direction (isotropic textures). The wavelet energy features of the directional textures change significantly when the image is rotated. However, for the isotropic images, the wavelet features are not sensitive to rotation. Therefore, for the directional textures, it is essential to calculate the wavelet features along a specific direction. In the proposed approach, the Radon transform is first employed to detect the principal direction of the texture. Then, the texture is rotated to place its principal direction at 0 degrees. A wavelet transform is applied to the rotated image to extract texture features. This approach provides a features space with small intraclass variability and, therefore, good separation between different classes. The performance of the method is evaluated using three texture sets. Experimental results show the superiority of the proposed approach compared with some existing methods.

Thalamus is an important neuro-anatomic structure in the brain. In this paper, an automated method is presented to segment thalamus from magnetic resonance images (MRI). The method is based on a discrete dynamic contour model that consists of vertices and edges connecting adjacent vertices. The model starts from an initial contour and deforms by external and internal forces. Internal forces are calculated from local geometry of the model and external forces are estimated from desired image features such as edges. However, thalamus has low contrast and discontinues edges on MRI, making external force estimation a challenge. The problem is solved using a new algorithm based on fuzzy C-means (FCM) unsupervised clustering, Prewitt edge-finding filter, and morphological operators. In addition, manual definition of the initial contour for the model makes the final segmentation operator-dependent. To eliminate this dependency, new methods are developed for generating the initial contour automatically. The proposed approaches are evaluated and validated by comparing automatic and radiologist's segmentation results and illustrating their agreement.

This paper presents a fast method for delineation of activated areas of the brain from functional magnetic resonance imaging (fMRI) time series data. The steps of the work accomplished are as follows. 1) It is shown that the detection performance evaluated by the area under the receiver operating characteristic curve is directly related to the signal-to-noise ratio (SNR) of the composite image generated in the detection process. 2) Detection and segmentation of activated areas are formulated in a vector space framework. In this formulation, a linear transformation (image combination method) is shown to be desirable to maximize the SNR of the activated areas subject to the constraint of removing inactive areas. 3) An analytical solution for the problem is found. 4) Image pixel vectors and expected time series pattern (signature) for inactive pixels are used to calculate weighting vector and identify activated regions. 5) Signatures of the activated regions are used to segment different activities. 6) Segmented images by the proposed method are compared with those generated by the conventional methods (correlation, t-statistic, and z statistic). Detection performance and SNRs of the images are compared. The proposed approach outperforms the conventional methods of fMRI analysis. In addition, it is model-independent and does not require a priori knowledge of the fMRI response to the paradigm. Since the method is linear and most of the work is done analytically, numerical implementation and execution of the method are much faster than the conventional methods.

A new method for automatic landmark extraction from MR brain images is presented. In this method, landmark extraction is accomplished by modifying growing neural gas (GNG), which is a neural-network-based cluster-seeking algorithm. Using modified GNG (MGNG) corresponding dominant points of contours extracted from two corresponding images are found. These contours are borders of segmented anatomical regions from brain images. The presented method is compared to: 1) the node splitting-merging Kohonen model and 2) the Teh-Chin algorithm (a well-known approach for dominant points extraction of ordered curves). It is shown that the proposed algorithm has lower distortion error, ability of extracting landmarks from two corresponding curves simultaneously, and also generates the best match according to five medical experts.

A new method is proposed for activation detection in event-related functional magnetic resonance imaging (fMRI). The method is based on the analysis of selected resolution levels (a subspace) in translation invariant wavelet transform (TIWT) domain. Using a priori knowledge about the activation signal and trends, we analyze their power in different resolution levels in TIWT domain and select an optimal set of resolution levels. A randomization-based statistical test is then applied in the wavelet domain for activation detection. This approach suppresses the effects of trends and enhances the detection sensitivity. In addition, since TIWT is insensitive to signal translations, the power analysis is robust with respect to signal shifts. The randomization test alleviates the need for assumptions about fMRI noise. The method has been applied to simulated and experimental fMRI datasets. Comparisons have been made between the results of the proposed method, a similar method in the time domain and the cross-correlation method. The proposed method has shown superior sensitivity compared to the other methods.

In this paper we present a novel method for the three-dimensional (3-D) centerline extraction of elongated objects such as vessels. This method combines the basic ideas in distance transform-based, thinning, and path planning methods to extract thin and connected centerlines. This efficient approach needs no user interaction or any prior knowledge of the object shape. We used the path planning approach, which has exclusively been used in the virtual endoscopy or robotics, to obtain the medial curve of the objects. To make our approach fully automated, a distance transform mapping is used to identify the end points of the object branches. The initial paths are also constructed on the surface of the object, traversing the same distance map. Then a thinning algorithm centralizes the paths. The proposed approach is especially efficient for centerline extraction of the complex branching structures. The method has been applied on the confocal microscopy images of rat brains and the results confirm its efficiency in extracting the medial curve of vessels, essential for the computation of quantitative parameters.

In this paper, we present a new curvature-based three-dimensional (3-D) deformable surface model. The model deforms under defined force terms. Internal forces are calculated from local model curvature, using a robust method by a least-squares error (LSE) approximation to the Dupin indicatrix. External forces are calculated by applying a step expansion and restoration filter (SEF) to the image data. A solution for one of the most common problems associated with deformable models, self-cutting, has been proposed in this work. We use a principal axis analysis and reslicing of the deformable model, followed by triangulation of the slices, to remedy self-cutting. We use vertex resampling, multiresolution deformation, and refinement of the mesh grid to improve the quality of the model deformation, which leads to better results. Examples of the model application to different cases (simulation, magnetic resonance imaging (MRI), computerized tomography (CT), and ultrasound images) are presented, showing diversity and flexibility of the model.

Background and Purpose: The purpose of this study was to show that the computer segmentation algorithm Iterative Self-Organizing Data Analysis Technique (ISODATA), which integrates multiple MRI parameters (diffusion-weighted imaging [DWI], T2-weighted imaging [T2WI], and T1-weighted imaging [T1WI]) into a single composite image, is capable of defining the ischemic lesion in a time-independent manner equally as well as the MRI techniques considered the best for each phase after stroke onset (ie, perfusion weighted imaging [PWI] and DWI for the acute phase and T2WI for the outcome phase)., Methods: We measured MRI parameters of PWI, DWI, T2WI, and T1WI from patients at the acute phase (<30 hours) and DWI, T2WI, and T1WI at the outcome phase (3 months) of ischemic stroke. The clinical neurological deficit was graded with the National Institutes of Health Stroke Scale (NIHSS). We compared the ISODATA lesion size with the PWI, DWI, and T2WI lesion sizes measured within the same slice at each phase. The lesion sizes were also correlated with NIHSS score of each phase., Results: We included 11 patients; 9 (82%) were women, and 7 (64%) were black. The mean+/-SD age was 65.5+/-9.3 years (range, 45 to 82 years). The median NIHSS score was 15 (minimum, 4; maximum, 24)at the acute phase and 3 (minimum, 0; maximum, 22) at the outcome phase. The median time interval from stroke symptom onset to the acute MRI study was 10 hours (range, 6 to 29 hours), and the mean time interval to the outcome study was 93+/-11 days (range, 72 to 106 days). In the acute phase, the ISODATA lesion size had high correlation with the PWI lesion size (r=0.95; 95% CI, 0.89 to 0.98; P<0.0001), DWI lesion size (r=0.83; 95% CI, 0.66 to 0.92; P<0.0001), and T2WI lesion size (r=0.67; 95% CI, 0.39 to 0.84; P=0.008) and moderate correlation with NIHSS score (r=0.59; 95% CI, 0.02 to 0.88; P=0.06). In the outcome phase, the ISODATA lesion size had high correlation with the T2WI lesion size (r=0.97; 95% CI, 0.94 to 0.99; P<0.0001) and NIHSS score (r=0.78; 95% CI, 0.34 to 0.94; P=0.004)., Conclusions: The integrated ISODATA method can identify and characterize the ischemic lesion independently of time elapsed since stroke onset. The ISODATA lesion size highly correlates with the PWI and DWI lesion size in the acute phase and with the T2WI lesion size in the outcome phase of ischemic stroke, as well as with the clinical neurological status of the patient.

Background and Purpose: After stroke, brain tissue undergoes time-dependent heterogeneous histopathological change. These tissue alterations have MRI characteristics that allow segmentation of ischemic from nonischemic tissue. Moreover, MRI segmentation generates different zones within the lesion that may reflect heterogeneity of tissue damage., Methods: A vector tissue signature model is presented that uses multiparametric MRI for segmentation and characterization of tissue. An objective (unsupervised) computer segmentation algorithm was incorporated into this model with the use of a modified version of the Iterative Self-Organizing Data Analysis Technique (ISODATA). The ability of the model to characterize ischemic tissue after permanent middle cerebral ischemia occlusion in the rat was tested. Multiparametric ISODATA measurements of the ischemic tissue were compared with quantitative histological characterization of the tissue from 4 hours to 1 week after stroke., Results: The ISODATA segmentation of tissue identified a gradation of cerebral tissue damage at all time points after stroke. The histological scoring of ischemic tissue from 4 hours to 1 week after stroke on all the animals was significantly correlated with ISODATA segmentation (r=0.78, P<0.001; n=20) when a multiparametric (T2-, T1-, diffusion-weighted imaging) data set was used, less correlated (r=0.70, P<0.01; n=20) when a T2- and T1-weighted data set was used, and not correlated (r=-0.12, P>0.47; n=20) when only a diffusion-weighted imaging data set was used., Conclusions: Our data indicate that an integrated set of MRI parameters can distinguish and stage ischemic tissue damage in an objective manner.

This study presents histological validation of an objective (unsupervised) computer segmentation algorithm, the iterative self-organizing data analysis technique (ISODATA), for analysis of multiparameter magnetic resonance imaging (MRI) data in experimental focal cerebral ischemia. T2-, T1-, and diffusion (DWI) weighted coronal images were acquired from 4 to 168 hours after stroke on separate groups of animals. Animals were killed immediately after MRI for histological analysis. MR images were coregistered/warped to histology. MRI lesion areas were defined using DWI, apparent diffusion coefficient (ADC) maps, T2-weighted images, and ISODATA. The last techniques clearly discriminated between ischemia-altered and morphologically intact tissue. ISODATA areas were congruent and significantly correlated (r = 0.99, P < 0.05) with histologically defined lesions. In contrast, DWI, ADC, and T2 lesion areas showed no significant correlation with histologically evaluated lesions until subacute time points. These data indicate that multiparameter ISODATA methodology can accurately detect and identify ischemic cell damage early and late after ischemia, with ISODATA outperforming ADC, DWI, and T2-weighted images in identification of ischemic lesions from 4 to 168 hours after stroke., (Copyright 2000 Wiley-Liss, Inc.)

The determination of volumes in clinical MRI studies are prohibitive because of the time required to compute an accurate volume. Techniques that speed up the calculation are prone to large errors which make most impractical for an accurate diagnosis. A linear filter, called the eigenimage filter, has been developed that separates a desired feature from other features which interfere with its observation in an image. Using the images produced by this technique (eigenimages), the amount of operator interaction required to calculate volumes are significantly reduced. The technique also has the ability to correct for partial volume averaging effects and as a result a more accurate volume can be determined. The technique was applied to a computer simulation and two phantom studies. The time required to calculate the volume was less than 1 min per slice and the errors in accuracy and reproducibility were less than 2% for all studies.

Background/Objective: White matter tractography is a non-invasive method, which reconstructs three-dimensional trajectories of the brain tracts using diffusion-tensor imaging (DTI) data. Due to the partial volume effect of DTI data, some of tractography algorithms are unable to follow the correct pathways after the crossing and branching regions. The main challenge for tractography methods has been the ability to detect these regions. Fast marching techniques are capable of tracking the fibers with wide spreading. Materials and Methods: In order to detect true fibers, an adaptive functional anisotropy (FA) weighted function is proposed to modify the speed function of these algorithms. The performance of the proposed tractography method is assessed using synthetic data and its feasibility is showed by extracting some well-known tracts using healthy human DTI datasets. Result: The percentage of the length of whole tracts extracted by our proposed method is above 85% even for a signal to noise ratio (SNR) level equal to 16. The ability of this method to detect the fiber crossing in simulation data is above 90%. Furthermore, the tractography results of some well-known tracts demonstrate the ability of the proposed methods to extract the correct pathways from the anatomical point of view. Conclusion: This method has led to great impact on the fast-marching fiber-tracking method in propagating the tractography front in an adaptive manner. The suggested speed function can make the speed of front propagation adapted to the type of brain's environments such as isotropic and anisotropic regions. [ABSTRACT FROM AUTHOR]

Abstract: Although Multiscale Cancer Modeling has a realistic view in the process of tumor growth, its numerical algorithm is time consuming. Therefore, it is problematic to run and to find the best treatment plan for chemotherapy, even in case of a small size of tissue. Using an artificial neural network, this paper simulates the multiscale cancer model faster than its numerical algorithm. In order to find the best treatment plan, it suggests applying a simpler avascular model called Gompertz. By using these proposed methods, multiscale cancer modeling may be extendable to chemotherapy for a realistic size of tissue. In order to simulate multiscale model, a hierarchical neural network called Nested Hierarchical Self Organizing Map (NHSOM) is used. The basis of the NHSOM is an enhanced version of SOM, with an adaptive vigilance parameter. Corresponding parameter and the overall bottom-up design guarantee the quality of clustering, and the embedded top-down architecture reduces computational complexity. Although by applying NHSOM, the process of simulation runs faster compared with that of the numerical algorithm, it is not possible to check a simple search space. As a result, a set containing the best treatment plans of a simpler model (Gompertz) is used. Additionally, it is assumed in this paper, that the distribution of drug in vessels has a linear relation with the blood flow rate. The technical advantage of this assumption is that by using a simple linear relation, a given diffusion of a drug dosage may be scaled to the desired one. By extracting a proper feature vector from the multiscale model and using NHSOM, applying the scaled-best treatment plans of Gompertz model is done for a small size of tissue. In addition, simulating the effect of stress reduction on normal tissue after chemotherapy is another advantage of using NHSOM, which is a kind of “emergent”. [Copyright &y& Elsevier]