Biomedical Image Processing / Medical Image Processing
Ali Kermani; Ahmad Ayatollahi; Sorour Mohajerani
Volume 8, Issue 4 , February 2015, , Pages 325-337
Abstract
IVUS imaging is a minimally invasive blood vessel cross-sectional imaging procedure in which accurate data is obtained from what is in there. Processing on these images or raw signals can provide wide range information for experts and practitioners, and can help them in making an accurate diagnosis and ...
Read More
IVUS imaging is a minimally invasive blood vessel cross-sectional imaging procedure in which accurate data is obtained from what is in there. Processing on these images or raw signals can provide wide range information for experts and practitioners, and can help them in making an accurate diagnosis and appropriate treatment. Extraction of tissue boundaries in the blood vessels is one of the challenging parts as a first step in this direction. In this paper a new method was proposed based on the minimax technique and connected components for extracting Adventitia tissue boundary in intravascular ultrasound images. For this purpose, initial boundary will be extracted using improved minimax technique. Then final boundary is extracted with high precision using connected components. The method was tested on a set of real data with regard to the Hausdorff distance and Jaccard index to evaluate its performance. Mean of Hausdorff distance and mean of Jaccard index were obtained 95% and 0.45 millimeter, consequently. These results show that the proposed method in this paper can extract Adventitia tissue boundaries more accurately than existing methods with regard to the distance Hausdorff distance and Jaccard index.
Biomedical Image Processing / Medical Image Processing
Sharare Kian-Bostanabad; Mahmoud Reza Azghani; Leila Rahnama
Volume 9, Issue 4 , February 2015, , Pages 341-350
Abstract
The cervical multifidus muscle is known as one of the deep neck extensor muscles that its dysfunction have been reported in people with neck pain.With regard to the limits on the evaluation of this muscle activity using electromyography, ultrasound was used to find out its function recently. The aim ...
Read More
The cervical multifidus muscle is known as one of the deep neck extensor muscles that its dysfunction have been reported in people with neck pain.With regard to the limits on the evaluation of this muscle activity using electromyography, ultrasound was used to find out its function recently. The aim of this study is evaluation of this muscle dimansions change during six shoulder joint activities in healthy subjects and people with chronic neck pain and providing predictive models. So The relationship between strength of shoulder joint during contraction with the changes of anterior-posterior dimension, lateral dimension, shape ratio and size of the cervical multifidus muscle were assessed using of Response Surface Method in the first step for subjects and activities and then for activities with subject blocking. Finally, predictive models were provided for abduction activity in 0-50% of maximum voluntary contraction (MVC) for healthy subjects and 50-100% for patients with data clustering. The anterior-posterior dimension showed a higher correlation with the shoulder joint strength than other factors. R2 values for this dimension in healthy subjects before and after data clustering is 0.552 and 0.66 and in patients is 0.339 and 0.505 respectively. Given the models correlation coefficient and its enhance by data clustering, it seems that evaluation of anterior-posterior dimension of this muscle during isometric abduction activiy of shoulder joint with the sttrength of 0-50% MVC for healthy subjects and 50-100% for patients with neck pain can be provide useful information about its function.
Biomedical Image Processing / Medical Image Processing
Mehdi Delavari; Amir Hosein Foruzan; Ben Vi Chen
Volume 8, Issue 3 , September 2014, , Pages 213-227
Abstract
Statistical Shape Models are used to interpret shapes. They include mean and variance of corresponding points of training shapes. One of the most important challenges in building statistical shape models is to establish correct correspondences among landmarks in a training set. In this paper, the ...
Read More
Statistical Shape Models are used to interpret shapes. They include mean and variance of corresponding points of training shapes. One of the most important challenges in building statistical shape models is to establish correct correspondences among landmarks in a training set. In this paper, the non-rigid CPD (Coherent Point Drift) method is used to find correct correspondences among points. This method uses both Deterministic Annealing and a non-rigid scheme to register two shapes simultaneously. Then, the statistical shape model is built using a rigid transformation. The proposed method is evaluated using Compactness, Generalization ability and Specificity measures. The built model is compared to models created using the ICP (Iterative Closest Point), TPS-RPM (Thin Plate Spline – Robust Point Matching) and MDL (Minimum Descreption Length) methods by these metrics. The results show that the proposed method performs like the MDL regarding Specificity measure (0.21±0.06). The Compactness and Generalization ability measures of the proposed method are very similar to those for the MDL method. The run-time of our proposed method is about 68 seconds which is faster than non-rigid TPS-RPM and MDL approaches (390 and 3600 seconds respectively). Our results are superior to the ICP and TPS-RPM algorithms.
Biomedical Image Processing / Medical Image Processing
Mahdie Ghasemi; Ali Mahloojifar; Mehdi Omidi
Volume 8, Issue 3 , September 2014, , Pages 261-275
Abstract
Functional changes in the brain motor network are responsible for the major clinical features of Parkinson’s disease (PD). Recent studies on investigation of the brain function show that there are spontaneous fluctuations between regions at rest as resting state network affected in various disorders. ...
Read More
Functional changes in the brain motor network are responsible for the major clinical features of Parkinson’s disease (PD). Recent studies on investigation of the brain function show that there are spontaneous fluctuations between regions at rest as resting state network affected in various disorders. In this paper, we examine changes of functional dependency between brain regions of interest associated with known anatomical pathology in Parkinson Disease (PD) using copula theory on resting state fMRI. Five types of copulas were tested: Gaussian and t (Euclidean), Clayton, Gumbel and Frank (Archimedean). We used an efficient maximum likelihood procedure for estimating copula parameters. Goodness of fits was tested using root mean square error (RMSE) and kulback-leibler divergence between each copula function and joint empirical cumulative distribution. Control vs PD group comparison was also done on dependency parameter using parametric and nonparametric tests. The results show that functional dependency between cerebellum and basal ganglia is much stronger in PD than in control. In this paper, we proposed for the first time that joint distribution characteristics could potentially provide information on discriminative features for functional connectivity analysis between healthy and patients.
Biomedical Image Processing / Medical Image Processing
Elahe Moghimirad; Ali Mahloojifar; Babak Mohammadzadeh Asl
Volume 8, Issue 3 , September 2014, , Pages 277-291
Abstract
A new implementation of a synthetic aperture focusing technique is presented in the paper. Standard medical ultrasound imaging is done using line-by-line transmission with classical Delay-and-Sum (DAS) image reconstruction. Synthetic aperture imaging, however, has a better resolution and frame rate in ...
Read More
A new implementation of a synthetic aperture focusing technique is presented in the paper. Standard medical ultrasound imaging is done using line-by-line transmission with classical Delay-and-Sum (DAS) image reconstruction. Synthetic aperture imaging, however, has a better resolution and frame rate in cost of more computational load. To overcome this problem, block processing algorithms are used in radar and sonar which are relatively unknown in medical. To extend the methods to medical field, one should concern the parameters difference such as carrier frequency, signal band width, beam width and depth of imaging. In this paper, we extended one of these methods called wavenumber to medical ultrasound imaging with a simple model of synthetic aperture focus. We have also used chirp pulse excitation followed by matched filtering, windowing and spotlighting algorithm to compensate the effect of differences in parameters between radar and medical imaging. Computational complexity of the two reconstruction methods, wavenumber and DAS, have been calculated. Field II simulated point data has been used to evaluate the results in terms of resolution and contrast. Evaluations with simulated data show that for typical phantoms, reconstruction by wavenumber algorithm is almost 20 times faster than classical DAS while retaining the resolution.
Biomedical Image Processing / Medical Image Processing
Leila Azimi; Nader Riahi Alam; Kavoos Firuozniya; Hamid Reza Saligheh Rad; Mojtaba Miri; Manizheh Pakravan; Anamollah Shahmohammadi
Volume 8, Issue 2 , June 2014, , Pages 113-123
Abstract
Gliomas are the most common primary neoplasm of the brain, varying histologically from low grade to high-grade. Perfusion-weighted MRI techniques have permitted the creation of cerebral blood volume (CBV) value, leading to the qualitative and quantitative assessment of tumor vascularity. This research ...
Read More
Gliomas are the most common primary neoplasm of the brain, varying histologically from low grade to high-grade. Perfusion-weighted MRI techniques have permitted the creation of cerebral blood volume (CBV) value, leading to the qualitative and quantitative assessment of tumor vascularity. This research aimed at assessing the rCBV and the ADC values in core and peritumoral areas glioma brain tumors and determining of significance rCBV of values i Alpha I In evaluating brain tumor. Ten patients with non-enhancing supratentorial gliomas were evaluated by diffusion weighted imaging (DWI) and standard dynamic susceptibility contrast-enhanced gradient echo during first pass of a bolus injection of contrast material before surgical resection. Six low-grade gliomas (WHO Grade II) and 4 high-grade gliomas (III, IV) were evaluated. Alpha Both the apparent diffusion coefficient (ADC) and relative cerebral blood volume (rCBV) values were calculated by a standard program on the solid portion of the tumor in the peritumoural area as well as contra lateral white matter. In perfusion, mean rCBVmax in high-grade and low-grade tumors were obtained (3.47±0.92), (2.37±0.49)-(1.66±0.68), (1.15±0.39) for core and peritumoral regions, respectively. In diffusion method, mean ADC in high-grade and low grade tumors were (0.53±0.07), (0.91±0.18)-(1.24±0.27), (1.007±0.33) for core and peritumoral regions, respectively. It was concluded that the values rCBVs are important in determining the grade of tumor and we propose that perfusion weighted imaging be done for all patients before surgery.
Biomedical Image Processing / Medical Image Processing
Sina Shamekhi; Mohammad Hossein Miranbaygi; Ali Gooya; Bahare Azarian
Volume 8, Issue 2 , June 2014, , Pages 183-202
Abstract
Two-dimensional gel electrophoresis (2DGE) is a basic and widely used method in proteomics. In this method, mixtures of proteins are separated due to the differences in their molecular weight and isoelectric points and a final image obtained from the separated protein spots is created. Due to the large ...
Read More
Two-dimensional gel electrophoresis (2DGE) is a basic and widely used method in proteomics. In this method, mixtures of proteins are separated due to the differences in their molecular weight and isoelectric points and a final image obtained from the separated protein spots is created. Due to the large number of the protein spots in a 2DGE image and the importance of separation of overlapping proteins, the image processing of these images is a complex process. 2DGE images pose various noises and artifacts such as cracks, staining artifacts, and streaks that affect the reliability of the analysis. In this work, we have proposed a novel spots filter based on the scale-space second order structural Hessian and its eigenvalues for enhancing and separating the spots from the background. Furthermore, in this work, 2DGE images have been segmented and the locations of the spots have been detected. To evaluate and compare the proposed method, we have implemented three methods: Otsu thresholding, Watershed transform, and the method proposed by Mylona et al. Based on the regional spot volume evaluation, the TPR and FPR of the proposed method are 78.6 and 14.9, the TPR and FPR of the Otsu method are equal to 71.4 and 25.7 percent, and the TPR and FPR of the Watershed algorithm are 53.9 and 8.1 percent, respectively. Also, in the spot counts evaluation, the Precision and TPR of the proposed method are equal to 83.6 and 81.1 percent, and the Precision and TPR of Otsu method are 65.4 and 78.3, respectively. The Watershed transform has detected the spots with Precision and TPR equal to 27.7 and 68.2 percent, and the Precision and TPR of the method proposed by Mylona et al. are 74.0 and 72.7 percent, respectively. The results reveal the accuracy and superiority of the proposed method.
Biomedical Image Processing / Medical Image Processing
Malihe Miri; Mohammad Taghi Sadeghi; Vahid Abootalebi
Volume 8, Issue 1 , March 2014, , Pages 45-56
Abstract
Successful outcomes of Sparse Representation-based Classifier (SRC) and Sparse Subspace Clustering (SSC) in many applications motivated us to combine these methods and propose a hierarchical classifier. The main idea behind the SRC and SSC algorithms is to represent a data using a sparse linear combination ...
Read More
Successful outcomes of Sparse Representation-based Classifier (SRC) and Sparse Subspace Clustering (SSC) in many applications motivated us to combine these methods and propose a hierarchical classifier. The main idea behind the SRC and SSC algorithms is to represent a data using a sparse linear combination of elementary signals so that those elementary signals which are similar to the data contribute mainly in the representation. In this paper, the performance of a sparse representation based classifier is improved by pre-clustering of training samples using the SSC algorithm. A twostage SRC is then designed using the resulting clusters. A test data is classified by first determining the most similar cluster. The data label is subsequently found using the second stage classifier. The performance of the proposed method is evaluated considering cancer classification problem using the 14-Tumors microarray dataset. Due to low number of data samples per each class and high dimensionality of the data, this is a challenging problem. Curse of dimensionality, overfitting of the classifier to the training data and computational complexity are the possible related problems. Our experimental results show that the proposed method outperforms some other state of the art classifiers.
Biomedical Image Processing / Medical Image Processing
Mohammad Reza Rezaeian; Gholam Ali Hossein-Zadeh; Hamid Soltanian Zadeh
Volume 8, Issue 1 , March 2014, , Pages 87-99
Abstract
Chemical exchange saturation transfer (CEST) is a new mechanism of contrast generation in magnetic resonance imaging (MRI) which differentiates molecule biomarkers via chemical shift. CEST MRI contrast mechanism is very complex and depends on radio frequency (RF) power and RF pulse shape. Two approaches ...
Read More
Chemical exchange saturation transfer (CEST) is a new mechanism of contrast generation in magnetic resonance imaging (MRI) which differentiates molecule biomarkers via chemical shift. CEST MRI contrast mechanism is very complex and depends on radio frequency (RF) power and RF pulse shape. Two approaches have been used to saturate contrast agent (CA) protons: continuous wave CEST (CW-CEST) and pulsed CEST. To find the optimal RF pulse, numerical solution of Bloch-McConnell equations (BME) may be used. In this paperwe find the optimum values of RF pulse parameters that maximize the CEST contrast. Discrete pulses have lower specific absorption ratio (SAR) than CW RF pulses. However, since discretization is performed on continuous RF pulses, optimizing the continuous RF pulses leads to the optimization of discrete RF pulses. Therefore, in this paper, Rectangular, Gaussian and Fermi pulses are investigated as CW RF pulses. In this investigation, in addition to considering the SAR limitation, 60 dB approximation for the RF pulse amplitude is used. To compare the efficiency of pulses, their resultant flip angles (FA) are assumed equal. Efficiency of CW-CEST is investigated using two parameters, CEST ratio and SAR. According to these parametres, rectangular, Fermi and Gaussian RF pulses have the best performance respectively. Since implementation of rectangular RF is harder than Gaussian and Fermi RF pulses, Fermi and Gaussian RF pulses are desired. Our results suggest that it is possible to maximize CEST ratio by optimizing parameters of rectangular (with an amplitude of 5.7μT), Gaussian (σ about 0.7s) and Fermi (a-value about 0.3s) pulses. Results are verified by empirical formulation of CEST ratio.
Biomedical Image Processing / Medical Image Processing
Nikta Jalayer; Majid Bagheri; Majid Pouladian
Volume 7, Issue 3 , June 2013, , Pages 209-217
Abstract
Recent developments in three-dimensional (3D) PET systems have enabled the spatial resolution to reach the 2- to 5-mm full-width-at-half-maximum (FWHM) range. With such improvements in spatial resolution, even small amounts of motion during PET imaging become a significant source of resolution degradation. ...
Read More
Recent developments in three-dimensional (3D) PET systems have enabled the spatial resolution to reach the 2- to 5-mm full-width-at-half-maximum (FWHM) range. With such improvements in spatial resolution, even small amounts of motion during PET imaging become a significant source of resolution degradation. In other words, increased spending on new-generation scanners can be fully justified only when appropriate motion correction methods are considered, to achieve the true resolution of the scanner. Motion correction methods developed for single photon emission CT (SPECT) are not necessarily applicable to PET because they may rely on the time-dependence of projections in SPECT (due to a rotating head or heads), which is not the case in PET. Nevertheless, a number of other methods implemented in SPECT are equally applicable to PET. In this work has been broadly categorized into the review and discussion of advanced correction methods for the cases of unwanted patient motion, motion due to cardiac cycles, and motion due to respiratory cycles. After reviewing some current methods, the model is introduced which was developed with the help of NCAT phantom and Sim SET. Two phantoms were extracted, male and female, from NCAT to see the differences between the results with the changes in the anatomy of these two phantoms. Then PET images were produced using Sim SET for all the phantoms available (with respiratory motion and without respiratory motion and for respiratory cycles of 4, 5 and 6 seconds for both male and female phantoms). The new model is introduced which is designed based on the respiratory cycle 5 seconds, using wavelet transforms. This model can track and compensate motion due to respiration. The results show that for the first frame and the last one because of very smooth and slight motions the images with motion are not that different from the images without motion, so the proposed model is not responding better than the images with motion. However, for the rest of the frames the model provides better images compare to the images with motion. Comparing to other methods, this model not only provides a good estimation for motion but also it doesn’t include the errors caused by markers and monitoring systems.
Biomedical Image Processing / Medical Image Processing
Neda Behzadfar; Hamid Soltanian Zadeh
Volume 7, Issue 3 , June 2013, , Pages 219-236
Abstract
Segmentation of tumors in magnetic resonance images is an important task. However, it is quite time consuming and has low accuracy and reproducibility when performed manually. Automating the process is challenging, due to high diversity in appearance of tumor tissue in different patients and in many ...
Read More
Segmentation of tumors in magnetic resonance images is an important task. However, it is quite time consuming and has low accuracy and reproducibility when performed manually. Automating the process is challenging, due to high diversity in appearance of tumor tissue in different patients and in many cases, similarity between tumor and normal tissues. This paper presents semi-automatic approach for analysis of multi-parametric magnetic resonance images (MRI) to segment a highly malignant brain tumor called Glioblastoma multiform (GBM). MRI studies of 12 patients with GBM tumors are used. To show that the proposed method identifies Gd-enhanced tumor pixels from T1-post contrast images minimal user interactions. They are also used to illustrate that the segmentation results obtained by the proposed approach are close to those of an expert, by showing excellent correlations among them (R2=0.97). In order to evaluate the proposed method in practical applications, effects of treatment of GBM brain tumors using Bevacizumab are predicted. Bevacizumab is a recent therapy for stopping tumor growth and even shrinking tumor through inhibition of vascular development (angiogenesis). To this end, two image series of 12 patients before and after treatment and relative changes in the volumes of the Gd-enhanced regions in T1-post contrast images are used as measure of response. The proposed method applies signal decomposition with KNN classifier to minimize user interactions and increase reproducibility of the results. Then histogram analysis is applied to extract statistical features from Gd-enhanced regions of tumor and quantify its micro structural characteristics. Predictive models developed in this work have large regression coefficients (maximum R2=0.91) indicating their capability to predict response to therapy. The results obtained by the proposed approach are compared with those of previous work where excellent correlations are obtained.
Biomedical Image Processing / Medical Image Processing
Maryam Afzali; Emadoddin Fatemizadeh; Hamid Soltanian Zadeh
Volume 7, Issue 1 , June 2013, , Pages 57-64
Abstract
Diffusion tensor magnetic resonance imaging (DTMRI) is a non-invasive method for investigating the brain white matter structure. It can be used to evaluate fiber bundles in the brain but in the regions with crossing fibers, it fails. To resolve this problem, high angular resolution diffusion imaging ...
Read More
Diffusion tensor magnetic resonance imaging (DTMRI) is a non-invasive method for investigating the brain white matter structure. It can be used to evaluate fiber bundles in the brain but in the regions with crossing fibers, it fails. To resolve this problem, high angular resolution diffusion imaging (HARDI) with a large number of diffusion encoding directions is used and for reconstruction, the Q-ball method is applied. In this method, orientation distribution function (ODF) of fibers can be calculated. Mathematical models play a crucial role in the field of ODF. For instance, in registering Q-ball images for applications like group analysis or atlas construction, one needs to interpolate ODFs. To this end, principal diffusion directions (PDDs) of each ODF are needed. In this paper, PDDs are defined as vectors that connect the corresponding local maxima of ODF values. Then, ODFs are interpolated using PDDs.We find the principal direction of ODF of the dataset to be interpolated and then rotate it to lie in the direction of the reference dataset. Now that ODFs are parallel, we apply linear interpolation to generate interpolated data. The proposed method is evaluated and compared with previous protocols. Experimental results show that the proposed interpolation algorithm preserves the principal direction of fiber tracts without producing any deviations in the tracts. It is shown that changes in the entropy of the interpolated ODFs are almost linear and the bloating effect (blurring of the principal directions) can be removed.
Biomedical Image Processing / Medical Image Processing
Pedram Masaeli; Hamid Behnam; Zahra Alizadeh Sani; Ahmad Shalbaf
Volume 7, Issue 3 , June 2013, , Pages 237-254
Abstract
Coronary artery diseases cause more than half of all deaths in the world. Obviously, early identification is an important way to control coronary artery disease that is diagnosed by measurement and scoring general and regional movement of left ventricle of heart (Normal, Hypokinetic and Akinetic). The ...
Read More
Coronary artery diseases cause more than half of all deaths in the world. Obviously, early identification is an important way to control coronary artery disease that is diagnosed by measurement and scoring general and regional movement of left ventricle of heart (Normal, Hypokinetic and Akinetic). The most common method of imaging the heart using ultrasound is called echocardiography. Using this method accurate view of the heart walls, valves and beginning of main arteries can be obtainbed. Due to the difficulty for the interpretation of these images, time consumption and errors in manual analysis methods, an automated analysis method is required. In this paper we calculate the displacement field in a cycle of heart motion from two-dimensional echocardiography images. To do this, a frame is usually chosen as the reference frame and then all images in a cycle are mapped to it with a mathematical equation. The main idea is to find a semi-local spatiotemporal parametric model for deformation created in a cardiac cycle with nonrigid registration using B-spline functions; as an optimization problem that effectively corrects differences due to movements by minimizing the difference between current frame and a reference frame. Motion estimation accuracy is measured using the sum of squares differences. We use gradient-descend algorithm and multiresolution method to acquire the coefficients in the motion model. The accuracy of the proposed method is assessed using a synthesis sequence of cardiac cycles produced with the simulation software Field II. This algorithm can be applied for the clinical analysis of regional left ventricle then movement parameters and threshold values for the scoring of each section can be extracted. The algorithm represents significant difference between a part of the normal heart and unhealthy heart that shows potential of clinical applications of the proposed method.
Biomedical Image Processing / Medical Image Processing
Marzie Ershad; Alireza Ahmadian; Houshang Saberi
Volume 7, Issue 2 , June 2013, , Pages 155-162
Abstract
Registration of preoperative images to intra-operative patient space is a crucial step in image guided surgery for tracking surgical tools relative to patient’s anatomy. In image guided spine surgery, due to the difference in patient’s positioning in preoperative imaging, compared with intra-operative ...
Read More
Registration of preoperative images to intra-operative patient space is a crucial step in image guided surgery for tracking surgical tools relative to patient’s anatomy. In image guided spine surgery, due to the difference in patient’s positioning in preoperative imaging, compared with intra-operative situation, there is a difference in spine curvature in these two positioning which means that a single rigid registration is not sufficient for registering the whole spine and it is necessary for each vertebra to be registered separately as a rigid body and with it’s appropriate transformation parameters. The registration was carried out using ICP algorithm. For evaluating the registration, TRE was calculated in the pedicle of the vertebra which is the target in pedicle screw insertion. In order to optimize the TRE this study was focused on the factors affecting TRE including different configuration of landmarks used in registration and the registration algorithm. Optimal configurations for the landmarks used in the registration were proposed and FLE for the point pairs were included in the registration algorithm to increase the registration accuracy. The results indicate a total improvement of 45% in the registration accuracy by optimizing the landmarks’ configuration and the registration algorithm.
Biomedical Image Processing / Medical Image Processing
Abbas Biniaz; Ataollah Abbasi; Mousa Shamsi
Volume 7, Issue 2 , June 2013, , Pages 175-186
Abstract
Segmentation divides an image to some subdivisions where which of ones has similar intensity gray levels. Among clustering methods fuzzy c-means (FCM) clustering has been frequently used for segmentation of medical images. However, this algorithm doesn’t incorporate spatial neighborhood information ...
Read More
Segmentation divides an image to some subdivisions where which of ones has similar intensity gray levels. Among clustering methods fuzzy c-means (FCM) clustering has been frequently used for segmentation of medical images. However, this algorithm doesn’t incorporate spatial neighborhood information in segmentation. This approach is very susceptible to nuisance factors. Therefore this paper proposes a Gaussian spatial FCM (gsFCM) to MR image segmentation. Proposed method has less sensitivity to noise specially in tissue boundaries, angles, and borders than spatial FCM (sFCM). Furthermore by the suggested algorithm a pixel which is a separate tissue from structurally point of view for example a tumor in primary stages of its appearance, has more chance to be a unique cluster. Applying quantitative assessments using Jaccard similarity index, Dice coefficient, and other validation functions on FCM,sFCM and gsFCM approaches show efficient performance of the proposed method. In this research the ISBR data bank is used for simmulations.Moreover in medical applications getting patient condition and information with fast methods is very important especially in emergency circumstances. Therefore all effective agents in patient health must be fast even medical algorithms such as clustering ones . Hence in this paper to decrease the time of convergence considerably and decline the number of iterations significantly, cluster centroids are initialized by an algorithm.
Biomechanics of Bone / Bone Biomechanics
Mina Gharenazifam; Ehsan Arbabi
Volume 6, Issue 4 , June 2012, , Pages 267-278
Abstract
One of the main causes of early osteoarthritis of the hip is Femoroacetabular Impingement (FAI). When the femoral head loses its spherical shape at head-neck junction, a special type of impingement, called Cam impingement, occurs. Alpha angle can be used as a geometric parameter for evaluating this kind ...
Read More
One of the main causes of early osteoarthritis of the hip is Femoroacetabular Impingement (FAI). When the femoral head loses its spherical shape at head-neck junction, a special type of impingement, called Cam impingement, occurs. Alpha angle can be used as a geometric parameter for evaluating this kind of anatomic deformity. In this article we propose a fully automatic strategy for estimating alpha angle by analyzing 3D data. In the proposed strategy a radial plane around the femoral head-neck axis is rotated in order to provide alpha angles in different orientations. For evaluating the proposed method, the alpha angle of twelve 3D femur models of female subjects, reconstructed from magnetic resonance images (including both right and left femur), have been estimated. The mean and standard deviation of these estimated alpha angles have been found to be in good agreement with the expected values for alpha angle in healthy human. In addition, the effect of the data resolution on the provided results has been evaluated in terms of accuracy and speed, by using four different resolutions of 3D meshes. The results indicate that using 64 times lower data resolution can increase the computational speed up to about 8 times and add an average error of about 2° to the estimated alpha angles.
Biomedical Image Processing / Medical Image Processing
Alireza Rahimpour; Abbas Nasiraei Moghaddam
Volume 6, Issue 3 , June 2012, , Pages 195-205
Abstract
Nowadays eye gaze tracking has wide range of applications in human computer interaction. One of these applications is using trajectory of eye gaze instead of foot or hand for disabled people to execute some commands. Various methods have been proposed, some of this methods can successfully track the ...
Read More
Nowadays eye gaze tracking has wide range of applications in human computer interaction. One of these applications is using trajectory of eye gaze instead of foot or hand for disabled people to execute some commands. Various methods have been proposed, some of this methods can successfully track the eye gaze. However, they always require specific circumstances, training or are not capable of real-time performance. In this paper, we proposed a framework to track eye gaze in real-time by using a simple and low cost webcam mounted on ordinary laptops. This process widely exploits the weighted normalized correlation function in an adaptive template matching approach. The implemented system tracks the face and also extracts some eye features such as iris position, eye corners and sclera region in eyes, in real time. These features are used in eye gaze estimation. Also the influence of illumination changes, background alterations, different faces and face movements is minimized as much as possible. The implemented gaze tracking system is able to control the motions of mouse cursor and click on an onscreen keyboard in real time.
Biomedical Image Processing / Medical Image Processing
Amin Mohammadian; Hasan Aghaeinia; Farzad Towhidkhah
Volume 6, Issue 3 , June 2012, , Pages 207-218
Abstract
In this paper, a method is proposed based on the prior knowledge from a new subject to improve the performance of person-independent facial expression recognition. First, in order to obtain a basic system, a combination of geometric features and texture descriptor is compared with global features (i.e., ...
Read More
In this paper, a method is proposed based on the prior knowledge from a new subject to improve the performance of person-independent facial expression recognition. First, in order to obtain a basic system, a combination of geometric features and texture descriptor is compared with global features (i.e., mapped face images using the Kernel-PCA and raw data of face images). The results of comparison under noisy conditions were investigated and evaluated by person-dependent/independent cross-validation method. The obtained basic system was evaluated by leave-one-subject-out cross-validation. Since the same subjects are not introduced in both training and test phases, the basic recognition system is person-independent and its performance is substantially lower than that of person-dependent cross-validation case. To improve the performance of the basic system, a method is proposed in which virtual samples are generated based on the prior knowledge from the new subject and are used in learning process. The results show that the recognition rate increases up to 96% for the person-dependent basic system, kernel-PCA method is more sensitive than the others to interpersonal variability, and the recognition rate is significantly (P<0.05) improved up to 91.39% compared to that of person-independent case.
Biomedical Image Processing / Medical Image Processing
Hadi Sabahi; Hamid Soltanian Zadeh; Lisa Scarpace; Tom Mikkelsen
Volume 5, Issue 4 , June 2011, , Pages 289-295
Abstract
In this paper, we propose a method to predict the outcome of Bevacizumab therapy on Glioblastoma Multiform (GBM) tumors. The method uses diffusion anisotropy indices (DAI) and spatial information to predict the treatment response of each tumor voxel. These DAIs are Fractional Anisotropy, Mean Diffusivity, ...
Read More
In this paper, we propose a method to predict the outcome of Bevacizumab therapy on Glioblastoma Multiform (GBM) tumors. The method uses diffusion anisotropy indices (DAI) and spatial information to predict the treatment response of each tumor voxel. These DAIs are Fractional Anisotropy, Mean Diffusivity, Relative Anisotropy, and Volume Ratio, extracted from Diffusion Tensor Imaging (DTI) data before treatment. The spatial information is considered as the distance of each tumor voxel from the tumor center, extracted from pre-treatment post-contrast T1-weighted Magnetic Resonance Images (pc-T1-MRI). DAIs and spatial information of each tumor voxel are considered as feature vector. DTI and pc-T1-MRI are gathered before and after the treatment of seven GBM patients. First, DAIs of all brain voxels and the distance of each tumor voxel from the tumor center are calculated. Second, the method registers pretreatment DAI maps and post-treatment pc-T1-MRI to pre-treatment pc-T1-MRI. Next, the tumor is segmented using thresholding technique from pc-T1-MRI. Then, Gd-enhanced voxels of the pre- and posttreatment pc-T1-MRI are compared to label the feature vectors. Three classifiers were evaluated, including Support Vector Machine, K-Nearest Neighbor, and Artificial Neural Network. Classification results show a preference for K-Nearest Neighbor based on well-established performance measures.
Biomedical Image Processing / Medical Image Processing
Fateme Bagheri; Hamid Behnam; Jahangir Tavakoli; Siavash Rahimian
Volume 5, Issue 2 , June 2011, , Pages 117-125
Abstract
In this study we evaluate parameter of nonlinearity and parameter of h by measuring of the amplitude of the second harmonic component and the fundamental component.This method is a variation of the finiteamplitude that has been adopted for pulse echo measurements. We used normal and cooked pork muscle ...
Read More
In this study we evaluate parameter of nonlinearity and parameter of h by measuring of the amplitude of the second harmonic component and the fundamental component.This method is a variation of the finiteamplitude that has been adopted for pulse echo measurements. We used normal and cooked pork muscle in-vitro. For B/A the result is showed as image and for h the result obtain as absolute mean.The result showed that these parameters can distinguish between normal and cooked tissue.This method was considered to be usable for control and monitoring HIFU.
Biomedical Image Processing / Medical Image Processing
Maryam Momeni; Hamid Abrishami Moghaddam; Reinhard Grebe; Kamran Kazemi; Fabrice Wallois
Volume 5, Issue 3 , June 2011, , Pages 231-244
Abstract
Reliable gradation of neonatal brain development is important for clinical investigation of neurological disorders. A prerequisite for such quantification of development is knowledge about an appropriate temporal resolvability. For this purpose, we investigated the evolution of macroscopic morphological ...
Read More
Reliable gradation of neonatal brain development is important for clinical investigation of neurological disorders. A prerequisite for such quantification of development is knowledge about an appropriate temporal resolvability. For this purpose, we investigated the evolution of macroscopic morphological features of the neonatal brain to estimate, for the first time, the required temporal interval in the early weeks after birth. In a first step, we constructed two neonatal templates for the age ranges of 39-40 and 41- 42 weeks' gestational age using T1-weighted MR images. We compared the spatial variation of anatomical landmarks and the average and the maximal length of spatial deformation in 25 subjects normalized to the two templates along x, y and z directions. MANOVA confirmed the significant difference between spatial variations of the above macroscopic features in the two age ranges. Furthermore, quantitative analysis of feature scattering yielded the same result even in features for which the null hypothesis was not rejected by MANOVA. We conclude that minimal temporal interval of two weeks is required for acute macroscopic morphological studies of the developing brain in the early weeks after birth.
Biomedical Image Processing / Medical Image Processing
Mohammad Hasan Moradi; Mohammad Sajad Manuchehri; Reza IraniRad
Volume 5, Issue 4 , June 2011, , Pages 313-331
Abstract
During the centuries, palpation has always been a crucial procedure in diagnosing the diseases. At first, these procedures were invasive, but nowadays numerous attempts by the name of elastographyhave been madeforreaching to noninvasive methods. Elastographys basic datais tissues relative displacement ...
Read More
During the centuries, palpation has always been a crucial procedure in diagnosing the diseases. At first, these procedures were invasive, but nowadays numerous attempts by the name of elastographyhave been madeforreaching to noninvasive methods. Elastographys basic datais tissues relative displacement which is tracked by ultrasound waves. First in these systems in order to attain the displacements gradient, an image of tissue is taken and then it is compared to image of that same tissue after applying a small mechanical impulse into it. Mechanical strain is calculated by estimating the displacements gradient and demonstrated as an image with gray levels named elastogram (strains image) .Based on how the mechanical vibration is given, ultrasound-elastography will separate into four categories as follows: static, dynamic, shear-wave and passive elastography. In static-elastography, the force is applied manually by the clinician and therefore it depends on operators skill and cannot be considerable. In dynamic type the movement of tissue is constantly provided by an external vibrator, so in order to prevent the interference of impulses we must use a rapid imaging system that eventually will cost extra expense and unavailability. Shear-wave elastography which currently is the most common method used in elastography systems,has an external vibratorLike dynamic method, but due to momentary impulses, it skips the problem of impulse interference. In passive method, physiologic movements of body will be given to tissue as itsvibration. This technique is hypothetical yet.
Biomedical Image Processing / Medical Image Processing
Fateme Nazem; Alireza Ahmadian; Mohammad Javad Abolhasani; Nasim Dadashi; Masoume Gity; Mohammad Bagher Shiran
Volume 5, Issue 4 , June 2011, , Pages 351-358
Abstract
Abstract: Image guided liver surgery based on intra-operative ultrasound images has received much attention in recent years. Using an efficient point-based registration method to improve both the accuracy and computational time for registration of pre-deformation CT liver images to post-deformation Ultrasound ...
Read More
Abstract: Image guided liver surgery based on intra-operative ultrasound images has received much attention in recent years. Using an efficient point-based registration method to improve both the accuracy and computational time for registration of pre-deformation CT liver images to post-deformation Ultrasound images is of great concern during surgical procedure. Although, Iterative Closest Point (ICP) algorithm is widely used in surface-based registration, its performance is strongly dependent on existence of noise and initial alignment. The registration technique based on the Unscented Kalman Filter (UKF) proposed recently can be a solution to overcome to noise and outliers on an incremental registration basis but it suffers from computational complexity. To overcome the limitations of ICP and UKF algorithms we proposed an incremental two-stage registration algorithm based on the combination of ICP and UKF algorithm to update the registration process based on arrival of intra-operative images. The two-stage algorithm is examined on phantom data sets. The results of phantom study confirm that the two-stage algorithm outperforms the accuracy of ICP and UKF by 23% and 13%, respectively and reduces the running time of UKF by 60%.
Biomedical Image Processing / Medical Image Processing
Effat Yahaghi; Yashar Nohi; Amir Movafeghi; Hamid Soltanian Zadeh
Volume 4, Issue 1 , June 2010, , Pages 1-11
Abstract
Magnetic resonance imaging (MRI) is a non-ionizing method for identification and evaluation of soft tissue lesions. Perfusion MRI evaluates soft tissues by measuring changes in magnetization of water molecules due to a contrast agent. To this end, concentration curves in the plasma and tissue are estimated ...
Read More
Magnetic resonance imaging (MRI) is a non-ionizing method for identification and evaluation of soft tissue lesions. Perfusion MRI evaluates soft tissues by measuring changes in magnetization of water molecules due to a contrast agent. To this end, concentration curves in the plasma and tissue are estimated by MRI and effective longitudinal relaxation time (T1eff) of the tissue was calculated. To interpret the results, the effects of water exchange on the effective longitudinal relaxation time should be studied. This work presents such a study in which the equations of two- and three-compartmental models of rat brain tissue are solved using Hion and Runge-Kutta numerical methods for different input functions and simulated by Monte Carlo method. Since the exchange of water and contrast agent among different tissue compartments is a diffusion phenomenon, Monte Carlo method is applicable. Results of the numerical methods were compared with those of Monte Carlo simulation. The results of the two methods were almost identical with a maximum relative difference of less than 1%. In this work, concentration of contrast agent in plasma is estimated from MRI of a rat brain tissue. This data is used in the Monte Carlo method to obtain T1eff and exchange rate constants. An advantage of our method is that T1eff is obtained from real data and not from the curve fitting method as commonly used. We derive concentration of contrast agent as a function of time in extravascular space for different constants (K). Then, the curves of simulated and real data were compared to obtain the exchange rate constant of each compartment. The results showed that K of an abnormal tissue was larger than that of the normal tissues. As such, this parameter may be used for diagnosis and treatment of the soft tissue diseases.
Biomedical Image Processing / Medical Image Processing
Maede Hadinia; Reza Jafari
Volume 4, Issue 4 , June 2010, , Pages 317-326
Abstract
This paper presents image reconstruction in Diffuse Optical Tomography (DOT) using a high-order finite element method. DOT is a non-invasive imaging modality for visualizing and continuously monitoring tissue and blood oxygenation levels in brain and breast. Image reconstruction in DOT leads to an inverse ...
Read More
This paper presents image reconstruction in Diffuse Optical Tomography (DOT) using a high-order finite element method. DOT is a non-invasive imaging modality for visualizing and continuously monitoring tissue and blood oxygenation levels in brain and breast. Image reconstruction in DOT leads to an inverse problem consisting of a forward problem and an iterative algorithm. The inverse problem in DOT systems is ill posed and depends on the accuracy of the forward problem. An accurate model, that describes the light transmission in tissue is required and can increase the spatial resolution. Using first order finite elements in the forward problem, numerical results are converged to the exact solution with increasing the number of elements. However, increasing the number of elements may cause a critical issue in the ill-posed inverse problem. This paper focuses on applying the high-order finite element method without increasing the number of elements, and image reconstruction is accomplished. The forward problem results are compared with analytical solutions. Images of absorbers reconstructed using this method are presented.