Biomedical Image Processing / Medical Image Processing
Maryam Ashoori; Reza Aghaizadeh Zoroofi; Mohammad Sadeghi
Volume 17, Issue 2 , September 2023, , Pages 130-140
Abstract
Currently, the rapid growth of the beauty industry, along with the development of intelligent models based on machine learning algorithms, has led to an increase in extensive research in this field. Rhinoplasty is one of the most common and demanding facial cosmetic surgeries because the nose is the ...
Read More
Currently, the rapid growth of the beauty industry, along with the development of intelligent models based on machine learning algorithms, has led to an increase in extensive research in this field. Rhinoplasty is one of the most common and demanding facial cosmetic surgeries because the nose is the most prominence element of the face, which has a great impact on its attractiveness. The purpose of this article is to present a machine learning-based framework for predicting nasal aesthetic evaluation. In this article, a set of geometric parameters of the nose relative to the entire face are given as input and human rating as output to the popular machine learning regression algorithms. An ablation study was then carried out to examine the influence of facial shape, skin color, and texture on the beauty of the nose. Multilayer perceptron classification, K-means clustering, and grey level co-occurrence matrix were used to extract facial shape, skin color, and texture. The results show that the model based on geometric parameters has a moderate correlation with human rating, and by adding each subset of the features of face shape, color, and skin texture, the correlation of the obtained model increases until a high degree of correlation is achieved. The results also show that the random forest algorithm has the best performance among other algorithms based on the evaluation criteria of absolute mean error, root mean square error, and Pearson correlation. The results of this study show that the proposed framework can be helpful in determining the beauty of the nose.
Biomedical Image Processing / Medical Image Processing
Kambiz Rahbar; Fatemeh Taheri
Volume 17, Issue 2 , September 2023, , Pages 161-170
Abstract
Lung cancer is caused by the irregular and uncontrolled growth of cancer cells in the lung tissue. Cancer cells find the ability to divide and increase in an irregular and uncoordinated manner. The result of this proliferation is the formation of a cancerous mass in the lung. Lung cancer can start in ...
Read More
Lung cancer is caused by the irregular and uncontrolled growth of cancer cells in the lung tissue. Cancer cells find the ability to divide and increase in an irregular and uncoordinated manner. The result of this proliferation is the formation of a cancerous mass in the lung. Lung cancer can start in different parts of the lung, such as the bronchi (the air tubes that connect to the lungs) or non-bronchial tissues, and quickly spread to other parts of the body. The precise understanding of the mechanism of lung cancer is still a complex issue and many researches are being conducted in this field. However, early diagnosis has an important impact on the disease treatment process. Therefore, in this research, the diagnosis and classification of this disease is discussed with the help of deep learning and learning transfer. In this regard, the pre-trained Alexnet network has been selected. During the process of transfer learning, the network for lung cancer detection is set on IQ-OTH/NCCD data in three categories: normal, benign and malignant. For this purpose, the last all-connection layer of the Alexnet network is removed and replaced by a new all-connection layer corresponding to the number of layers in the dataset. The classification accuracy of the proposed method on the IQ-OTH/NCCD dataset is reported to be 93%.
Biomedical Image Processing / Medical Image Processing
Mohammad Mahdi Alimoradi; Mohammad Bagher Khodabakhshi; Shahriar Jamasb
Volume 17, Issue 1 , May 2023, , Pages 61-70
Abstract
Stroke is one of the causes of death and the main cause of disability in developed countries. Normally, identification of stroke lesions is done by magnetic imaging, and its analysis requires the continuous presence of a doctor in the treatment center. Therefore, intelligent processing of medical images ...
Read More
Stroke is one of the causes of death and the main cause of disability in developed countries. Normally, identification of stroke lesions is done by magnetic imaging, and its analysis requires the continuous presence of a doctor in the treatment center. Therefore, intelligent processing of medical images will be an effective approach for automatic diagnosis of brain lesions.In this paper, a new integrated framework based on fuzzy inference system and deep neural network for automatic segmentation of brain lesions is introduced. In this regard, firstly, an improved U-net deep network (U-net) has been introduced for lesion detection and segmentation, which includes increasing the number of encoder and decoder layers along with changing the activation functions. Then, by using a fuzzy inference system based on if-then rules used by membership functions, the proposed approach of this study, which is based on the pre-processing of input images and the use of the unit network, has been introduced.The results showed that the integration of the fuzzy inference system in the pre-processing with the improved deep network could increase the DICE coefficient up to 0.84. In addition, improving the contrast of the input images by the fuzzy system compared to the usual pre-processing methods such as histogram equalization showed a much better performance in the detection of lesions with small dimensions, which is due to the ability to control the amount of contrast increase in the fuzzy systems compared to the usual methods.
Biomedical Image Processing / Medical Image Processing
Mohamad reza Rezaeian
Volume 16, Issue 4 , March 2023, , Pages 300-310
Abstract
The chemical exchange due to saturation transfer by applying an electromagnetic radio frequency (RF) pulse to a magnetic resonance scanner is called the CEST effect. The CEST effect depends mainly on relaxation times, chemical exchange rate, concentration of the contrast agent and RF pulse properties. ...
Read More
The chemical exchange due to saturation transfer by applying an electromagnetic radio frequency (RF) pulse to a magnetic resonance scanner is called the CEST effect. The CEST effect depends mainly on relaxation times, chemical exchange rate, concentration of the contrast agent and RF pulse properties. Dependence of chemical exchange rate on some clinical indicators such as pH, temperature and glucose consumption, allows diagnosis diseases non-invasively. The chemical exchange rate is determined through presenting new objective function of the CEST effect in the mathematical closed form quantitatively. A new description of the optimal amplitude of the rectangular RF pulse is obtained by applying gradient-based methods on the proposed convex objective function. Chemical exchange rate is proposed at the simple representation form independent to contrast agent by reversing the optimal amplitude description for large shift frequency contrast agents. Evaluation of the objective function and the proposed relations are performed by comparing them with valid methods derived solving Bloch-McConnell equations through parametric and real data. The mean relative square error of the objective function based on the parametric data is 7.25% and for the proposed the optimal amplitude and chemical exchange rate based on the real data are 6.3% and 4.2%, respectively.
Biomedical Image Processing / Medical Image Processing
Sina Shamekhi
Volume 16, Issue 2 , September 2022, , Pages 95-113
Abstract
Intuitive examination of retinal layers in Spectral-Domain Optical Coherence Tomography (SD-OCT) images is one of the main methods used by physicians to diagnose retinal diseases. This method faces challenges such as noise and image complexity and the proximity of retinal layers. In recent years, the ...
Read More
Intuitive examination of retinal layers in Spectral-Domain Optical Coherence Tomography (SD-OCT) images is one of the main methods used by physicians to diagnose retinal diseases. This method faces challenges such as noise and image complexity and the proximity of retinal layers. In recent years, the automatic diagnosis of retinal diseases has become an important clinical issue in computer vision. In this research, a new method for efficient multi-class automatic classification of SD-OCT images has been proposed. This method consists of five stages, preprocessing, layer recognition, feature extraction, and image classification. Examination of the shape of the RNFL layer and IS/OS junction as a clinical method is influential in physicians' decisions to diagnose retinal diseases. Therefore, in this study, inspired by this clinical diagnosis method, the RNFL layer, and the IS/OS junction have been detected by a new method based on the Frangi vessel enhancement algorithm and the gradient of the image. Then, by extracting and selecting several efficient features from the curves of the layers, an algorithm based on the ensemble decision tree has been proposed for classifying SD-OCT images of the retina and presented as a MATLAB application. The proposed method has been evaluated using images of two well-known databases of Duke and Kermany. Based on the results, precision, sensitivity, specificity, accuracy, miss rate and F1-score of the proposed method in Duke database were equal to 98.7, 98.8, 99.4, 99.1, 1.3, and 98.7, respectively, and in Kermany database were 96.8, 96.7, 98.9, 98.4, 3.2 and 96.7 respectively. The results show the superiority of the proposed method compared to other comparative methods. In summary, the use of efficient features of retinal effective layers and a powerful algorithm for classification has improved the performance of the proposed method compared to previous more complex methods.
Biomedical Image Processing / Medical Image Processing
Maryam Dorvashi; Neda Behzadfar
Volume 15, Issue 4 , March 2022, , Pages 289-298
Abstract
Early detection of fatigue helps to improve the quality and effectiveness of neurofeedback training. Diagnosis of fatigue using the EEG signal of participants during neurofeedback training in 10 training sessions is reviewed in this paper. Neurofeedback training has two different neurofeedback training ...
Read More
Early detection of fatigue helps to improve the quality and effectiveness of neurofeedback training. Diagnosis of fatigue using the EEG signal of participants during neurofeedback training in 10 training sessions is reviewed in this paper. Neurofeedback training has two different neurofeedback training protocols called protocols one and two. The first protocol is a training feature, a combination of frequency and non-frequency features, but the second protocol only includes frequency features. In the first fatigue time protocol, the slope trend of the power changes of the second low alpha sub-band in the OZ channel is decreasing and the permutation entropy in the FZ channel is increasing. The slope of the score changes is also decreasing. In the second protocol, the slope trend of power changes is the second low alpha sub-band in the OZ channel and decreases the score, in other words, the lack of feature change in line with the goal of neurofeedback training is due to fatigue and the participant cannot score. The results are based on the power slope trend of the second lower alpha sub-band and permutation entropy, which indicates that fatigue occurs for one participant in the first protocol and for three participants in the second protocol.
Biomedical Image Processing / Medical Image Processing
Mohamad Reza Rezaeian
Volume 15, Issue 1 , May 2021, , Pages 47-58
Abstract
Molecular magnetic resonance imaging by tracking contrast agents based on magnetic resonance of the nucleus is considered a novel anatomical and functional diagnostic method in various medical applications due to its good spatial resolution and safe technology. In a magnetic resonance scanner, a spectroscopic ...
Read More
Molecular magnetic resonance imaging by tracking contrast agents based on magnetic resonance of the nucleus is considered a novel anatomical and functional diagnostic method in various medical applications due to its good spatial resolution and safe technology. In a magnetic resonance scanner, a spectroscopic spectrum known as the Z-spectrum is obtained by applying a predominantly rectangular electromagnetic saturation pulse. At frequencies corresponding to the Larmor frequency, some amplitudes due to water saturation contrast factors are formed, representing saturation transfer’s effect due to chemical exchange (CEST). Chemical shifts, magnetic field heterogeneity and imaging process’s noise, while shifting the Larmore frequencies position, distorts the CEST effect. This noise is mainly modeled by the raisin distribution, which is an extent of Gaussian distribution. In this paper, an efficient method for reducing noise from the Z-spectrum and detecting the CEST effect is presented. Deionization is performed using the analytical model’s output resulting from solving the Bloch-McCannell equations and detecting the CEST effect by calculating the Bayesian likelihood function. The proposed method’s effectiveness for noise cancellation and detection the CEST effect was performed on real Z-spectra which is obtained from magnetic resonance scanners and data obtained from human tissue. The average performance of the proposed method is measured by relative mean square error between the real Z-spectrum and the noise in the signal to noise 10dB and the number of observations 5 was about four percent. The value of the first type of error (p-value) based on parametric data was less than 5% when the noise variance was more than 0.008 and the number of observations was more than 5. In this paper, a criterion for detecting the effect of CEST based on the mediation operator is proposed to evaluate the efficiency of the proposed method in proportion to the noise power and the number of observations.
Biomedical Image Processing / Medical Image Processing
Dorsa Jafarkhah Seighalani; Mehran Yazdi; Mohammad Faghihi
Volume 14, Issue 4 , February 2021, , Pages 267-276
Abstract
Cancer is one of the most common diseases at the present time. Among different types of this disease, brain cancer has a high fatality rate and accurate and timely diagnosis of it, can have a major impact on the patient’s life. Doctors need MRI and CT scan of brain to diagnose this condition. A ...
Read More
Cancer is one of the most common diseases at the present time. Among different types of this disease, brain cancer has a high fatality rate and accurate and timely diagnosis of it, can have a major impact on the patient’s life. Doctors need MRI and CT scan of brain to diagnose this condition. A precise image processing technique can help the medical specialists and speed up the diagnosis process. Many methods have been proposed to recognize brain tumors in medical images; however their accuracies were not acceptable. In fact, low accuracy is a result of the similarities between brain and tumor tissue. In this paper we propose a tumor recognition method using fusion of MRI and CT Scan images. This method exploits a deep learning based feature extraction algorithm that helps to distinguish tumors from brain tissue. Tumor recognition and accuracy calculation is performed for three common types of brain tumors (glioma, meningioma, and pituitary tumor). Our results show a great improvement of performance in comparison to related works.
Biomedical Image Processing / Medical Image Processing
Gelareh Valizadeh; Farshid Babapour Mofrad; Ahmad Shalbaf
Volume 14, Issue 4 , February 2021, , Pages 291-306
Abstract
Statistical Shape Modeling is widely used in many applications of cardiac images. Many efforts have been done to generate optimized Statistical Shape Models (SSMs). In this paper, we evaluated three different 3D endocardial models constructed using different alignment procedures. From 20 healthy CMR ...
Read More
Statistical Shape Modeling is widely used in many applications of cardiac images. Many efforts have been done to generate optimized Statistical Shape Models (SSMs). In this paper, we evaluated three different 3D endocardial models constructed using different alignment procedures. From 20 healthy CMR datasets, three different endocardial models are generated by varying the surface alignment methods means based on the Center of the Apex (CoA), the Center of Mass (CoM), and the Center of the Basal (CoB) of the endocardium. Then Principle Component Analysis (PCA) is applied to show the maximum variation of the SSMs. The constructed statistical models are compared by measuring the compactness, generalization ability, and specificity. Besides, the performance of each model in the 3D endocardial segmentation application using the Active Shape Model (ASM) technique is evaluated by the Hausdorff Distance (HD) criterion. The results indicate that the CoB-based model is less compact than the CoA-based model but more compact than the CoM-based model. Although for a constant number of modes the reconstruction error is approximately the same for all models, surface alignment based on CoB leads to generate a more specific model than the others. The resulted HDs show that the CoB alignment strategy produces the ASM which has the best performance in 3D endocardial segmentation among the other models. The computed results from the quantitative analysis demonstrate that varying alignment strategies affect the quality of the constructed SSM. It is obvious that the specificity and segmentation accuracy of the proposed CoB-based model outperforms the classical CoM-based approach.
Biomedical Image Processing / Medical Image Processing
Amirhossein Chalechale; Ali Khadem
Volume 14, Issue 1 , May 2020, , Pages 31-42
Abstract
The well-timed and correct diagnosis of Bipolar Disorder (BD) followed by proper treatment is vital for avoiding the progress of the illness. Although using resting-state functional magnetic resonance imaging (rs-fMRI) data and the features extracted from them may have an important role in diagnosing ...
Read More
The well-timed and correct diagnosis of Bipolar Disorder (BD) followed by proper treatment is vital for avoiding the progress of the illness. Although using resting-state functional magnetic resonance imaging (rs-fMRI) data and the features extracted from them may have an important role in diagnosing this kind of brain disorder, few researches have been conducted on this illness and the obtained results are not accurate. In this research we used a new approach to diagnose BD I. By using seed-based correlation we used the following 4 regions of interest in order to extract the connectivity maps for each subject: the posterior cingulate cortex (PCC) to probe the default mode network (DMN), the amygdala and the subgenual cingulate cortex (sgACC) to probe the salience network (SN) and the dorsolateral prefrontal cortex (dlPFC) to probe the frontoparietal network (FPN). After computing the connectivity maps for each subject we extracted the most important connectivities using different threshold on the t-value from the t-test that we applied on them and then we used a support vector machine (SVM) using only four combined features and a leave one out cross-validation (LOOCV) method to classify the two groups. The proposed method was done on rs-fMRI data from 49 healthy control subjects and 34 BD I patients and an accuracy of higher than 90% was obtained in differentiating the two groups from each other. Also there were no hyper-connectivity between the 4 ROIs and the rest of the brain regions for the BD I groups in relation with the healthy controls. The regions that had most of the hypo-connectivity with the 4 ROI’s that we used were: the angular gyrus (Ag) and the orbitofrontal cortex (OFC) with the PCC, the anterior cingulate cortex with the amygdala and the dlPFC and the inferior temporal gyrus (ITG) with the sgACC.
Biomedical Image Processing / Medical Image Processing
Saeid Shakeri; Farnaz Ghassemi; Farshad Almasganj
Volume 13, Issue 1 , April 2019, , Pages 17-30
Abstract
Noise removal is one of the most important steps in digital image processing. Cone beam computed tomography (CBCT) is increasingly utilized in maxillofacial and dental imaging. Compared to conventional CT, CBCT images have diffrent noise and artifacts due to much less applied dose and their reconstruction ...
Read More
Noise removal is one of the most important steps in digital image processing. Cone beam computed tomography (CBCT) is increasingly utilized in maxillofacial and dental imaging. Compared to conventional CT, CBCT images have diffrent noise and artifacts due to much less applied dose and their reconstruction algorithm. Therefore, the use of noise reduction techniques in these images is necessary to increase the signal-to-noise ratio. In this paper, the independent component analysis (ICA) method has been used to seperate noise from CBCT images and three different ICA algorithms, NG-FICA, ERICA and FastICA were investigated. In addition, two powerful noise reduction method, 2D discrete wavelet thresholding and optimized anisotropic diffusion filter is used to evaluate the results. Our proposed method has been validated on 12 different images in the presence of Gaussian and Spectral noise and the results are evaluated using processing time criteria, PSNR, MSE and SSIM. The results show that the ICA methods have advantage in noise reduction from CBCT images compared to the other noise reduction methods and among the three studied ICA algorithms, the NG-FICA algorithm has better performance in terms of processing time, preserving image quality and noise reduction.
Biomedical Image Processing / Medical Image Processing
Sina Shamekhi; Mohammad Hosein Miranbeigi; Ali Gooya
Volume 12, Issue 4 , January 2019, , Pages 265-275
Abstract
Matching of the protein spots in two dimensional gel electrophoresis (2DGE) images is a main process of analyzing these images. Due to the challenges of 2DGE images such as the presence of noise and artifacts, the matching of protein spots is performed under human supervision. This supervision involves ...
Read More
Matching of the protein spots in two dimensional gel electrophoresis (2DGE) images is a main process of analyzing these images. Due to the challenges of 2DGE images such as the presence of noise and artifacts, the matching of protein spots is performed under human supervision. This supervision involves human errors. Therefore, in this work a new automated algorithm has been proposed for spot matching in 2DGE images which is based on a probabilistic model. Due to the complexities of the proposed model, the Variational Bayes has been used to solve the equations of the model. The performance of the proposed algorithm has been evaluated on real and synthetic 2DGE images with some statistical criteria. Protein spots in real image dataset have been matched by the proposed method with angular error of 0.05 and end point error of 1.46 and in synthetic image dataset with angular error of 0.13 and end point error of 0.90. These results reveal higher precision and effectiveness and lower matching error of the proposed method.
Biomedical Image Processing / Medical Image Processing
Abbas Biniaz; Fatemeh Abdolali; Reza Aghaeizadeh Zoroofi; Omid Haji Maghsoudi; Yoshinobu Sato
Volume 12, Issue 4 , January 2019, , Pages 317-329
Abstract
Wireless capsule endoscopy is a non-invasive diagnosis method which allows recording a video as the capsule travels through the gastrointestinal tract. The practical drawback is producing a long clinical video up to 8 hours and it takes about 2 hours to review the exam by an experienced expert. Video ...
Read More
Wireless capsule endoscopy is a non-invasive diagnosis method which allows recording a video as the capsule travels through the gastrointestinal tract. The practical drawback is producing a long clinical video up to 8 hours and it takes about 2 hours to review the exam by an experienced expert. Video summarization methods can reduce the time required by experts and errors in manual interpretation. This paper presents an automatic method based on unique properties of adaptive singular value decomposition through sliding window that can reduce the long annotation time. By utilizing these properties, we are able to summarize a WCE video by outputting a motion video summary. Moreover, we apply an effective approach based on adaptive contrast diffusion to correct uneven illumination that deal with the low contrast generally caused by poor visibility conditions of the GI tract, WCE power and its structure. Experimental results on WCE videos indicate that a significant reduction in the review time is feasible. Quantitative and qualitative results of summarization show the effectiveness of proposed method that can be adapted to various clinical applications, such as training of young physicians, computer assisted diagnosis, medical decision support or medical document management.
Biomedical Image Processing / Medical Image Processing
Amir Sezavar; Hassan Farsi; Farima Farsi
Volume 12, Issue 4 , January 2019, , Pages 341-355
Abstract
Prostate cancer is one of the most important diseases of men whose growth can be disrupted by early diagnosis of it. In order to determine the grade of prostate cancer, the biopsy is used and structure of tissue is examined under microscopes. According to new grading system, the prostate tissues are ...
Read More
Prostate cancer is one of the most important diseases of men whose growth can be disrupted by early diagnosis of it. In order to determine the grade of prostate cancer, the biopsy is used and structure of tissue is examined under microscopes. According to new grading system, the prostate tissues are grading to five categories, between 1 to 5, where the highest grade shows the worst condition. Since human grading is time consuming, automatic grading systems have been used since recent years. Although some efficient algorithms have been introduced for image classification, the semantic gap between low-level features and human visual concept is still an important reason not to achieve high precision. In this paper, a new method for prostate cancer grading is presented which uses a combination of deep features, extracted by convolional neural network (CNN), and stochastic tissue features, extracted using multi-level gray level co-occurrence matrixes (ML-GLCM). Therefore, high-level features are achieved by using CNN and by combining with stochastic tissue features, the grading precision is increased. In order to evaluate the proposed method, it is examined on the pathology prostate image database which is generated by international society of urological pathology (ISUP). Experimental results demonstrate that the proposed method achieves more accuracy than state-of-the-art methods on prostate cancer grading.
Biomedical Image Processing / Medical Image Processing
Somayeh Maleki Balajoo; Davoud Asemani; Hamid Soltanian-Zadeh
Volume 12, Issue 2 , September 2018, , Pages 111-124
Abstract
Early alterations of functional connectivity (FC) within the default mode network (DMN) have been reported in Alzheimer’s disease (AD). Recently, the resting-state brain networks have been described with non-stationary profiles since inter- and intra-FC of the brain networks changes over time, ...
Read More
Early alterations of functional connectivity (FC) within the default mode network (DMN) have been reported in Alzheimer’s disease (AD). Recently, the resting-state brain networks have been described with non-stationary profiles since inter- and intra-FC of the brain networks changes over time, even at rest. To fully understand the FC changes that characterize AD, the underlying change of its dynamic pattern needs to be captured. The purpose of this study was to evaluate dynamic FC (dFC) patterns within the DMN in patients with AD relative to healthy aging. Here, a sparse logistic regression (SLR) model was employed to estimate the dFC networks in patients with AD (n = 24) compared with healthy control group (n = 37) using resting-state functional magnetic resonance imaging (rs-fMRI) data. To analyze the dFC network, we introduced a temporal variability-functional pattern (TV-FP) score, which shows how the functional pattern of a given region changes over time. This score is able to quantify the temporal patterns of regions involved in a dFC network. We compared TV-FP score across groups. The results indicate that the main regions of DMN, such as the anterior medial prefrontal cortex (aMPFC) and lateral temporal cortex (LTC), are associated with a significantly increased TV-FP score in the AD group when compared to the HC group. The FC pattern of these regions is impaired in AD according to a conventional static functional connectivity (sFC) analysis. These findings may suggest that aMPFC and LTC may tend to reorganize their functional pattern to compensate for the related functional deficiency due to AD.
Biomedical Image Processing / Medical Image Processing
Hamed Fayyaz; Mohsen Soryani; Ehsan Koozegar; Tao Tan
Volume 12, Issue 2 , September 2018, , Pages 137-146
Abstract
Automated 3D breast ultrasound (ABUS) is a novel system for breast screening. It has been proposed as a supplementary modality to mammography for detection and diagnosis of breast cancers. Although ABUS has better performance for dense breasts, reading ABUS images is time-consuming ...
Read More
Automated 3D breast ultrasound (ABUS) is a novel system for breast screening. It has been proposed as a supplementary modality to mammography for detection and diagnosis of breast cancers. Although ABUS has better performance for dense breasts, reading ABUS images is time-consuming and exhausting. A computer-aided detection (CAD) system can be helpful for interpretation of ABUS images. Mass Segmentation in CADe and CADx systems play the leading role because it affects the performance of succeeding stages. Besides, it is a very challenging task because of the vast variety in size, shape, and texture of masses. Moreover, imbalanced datasets make segmentation harder. A novel mass segmentation approach based on deep learning is introduced in this paper. The deep network that is used in this study for image segmentation is inspired by U-net which has been used broadly for dense segmentation in recent years. Performance was determined using a dataset of 50 masses including 38 malignant and 12 benign masses.
Biomedical Image Processing / Medical Image Processing
Mahdieh Ghasemi
Volume 12, Issue 1 , June 2018, , Pages 51-61
Abstract
Parkinson’s disease (PD) is a progressive neurological disorder characterized by tremor, rigidity, and slowness of movements. Different pathological attacks in Parkinson’s disease can be investigated by directional relations in the base spontaneous fluctuations of the brain from the resting ...
Read More
Parkinson’s disease (PD) is a progressive neurological disorder characterized by tremor, rigidity, and slowness of movements. Different pathological attacks in Parkinson’s disease can be investigated by directional relations in the base spontaneous fluctuations of the brain from the resting state functional magnetic resonance imaging (RS-fMRI) data. In this paper, for analyzing the directional brain network at rest, Directed Transform Function (DTF) technique with graph theory has been used in two frequency sub-bands and intra/inter group connectivities were compared by statistical analysis. The result of group comparison between PD and healthy which has been done, showed that there are more significant connections in the low frequency band in Parkinson’s disease and control group compared to high frequency band. The relation between basal ganglia and cerebellum has been disturebed in Parkinson’s disease. Furthermore, some brain regions such as left cerebellum has the most information flow in healthy group which characterized by pivotal regions which were influenced by the other brain regions, this connection became disordered in Parkinsonism.
Biomedical Image Processing / Medical Image Processing
Seyed Hani Hojjati; Ataollah Ebrahimzadeh; Ali Khazaee; Abbas Babajani-Fermi
Volume 11, Issue 1 , May 2017, , Pages 29-40
Abstract
Predicting AD based on Brain network analysis has been the subject of much investigation. Here, we aim to identify the changes in brain in patients that conversion from (Mild Cognitive Impariment) MCI to AD (MCI-C) and non conversion from MCI to AD (MCI-NC), to provide an algorithm for classification ...
Read More
Predicting AD based on Brain network analysis has been the subject of much investigation. Here, we aim to identify the changes in brain in patients that conversion from (Mild Cognitive Impariment) MCI to AD (MCI-C) and non conversion from MCI to AD (MCI-NC), to provide an algorithm for classification of these patients by using a graph theoretical approach. In this algorithm we proposed Discriminant Correlation Analysis (DCA), feature level fusion for multimodal biometric recognition method were applied to the original feature sets. An accuracy of 86/167% was achieved for predicting AD using the DCA and the support vector machine classifier. We also performed a hub node analysis and found the number of hubs in progressive AD patients. Indeed, this is the first neuroimaging study that integrates rs-fMRI with sMRI for detecting conversion from MCI to AD. The proposed classification method highlights the potential of using both resting state fMRI and MRI data for identification of the early stage of AD.
Biomedical Image Processing / Medical Image Processing
Marjan Iranianpour Haghighi; Seyyed Vahab Shojaeddini
Volume 10, Issue 4 , January 2017, , Pages 291-302
Abstract
Detecting lesion borders is the first step for intelligent lesion identification in dermoscopy, therefore it may influence the accuracy and validity of the next steps of this process. Unfortunately, extracting borders is hampered by some challenges such as losses associated with irregular borders, poor ...
Read More
Detecting lesion borders is the first step for intelligent lesion identification in dermoscopy, therefore it may influence the accuracy and validity of the next steps of this process. Unfortunately, extracting borders is hampered by some challenges such as losses associated with irregular borders, poor contrast, and artifacts encountered in some area. In this paper, the improved version of energy function optimization technique is introduced in order to separate the skin and lesions in the processing of dermoscopy images. This technique is based on the concept of radial directions in the contour development process, which reduces the sensitivity of estimating the boundaries of lesions to the above constraints. The performance of the proposed method is evaluated on a dataset of dermoscopy images which are captured from various lesions with different sizes and boundaries. The obtained results of the proposed method are compared with some other state-of-the-art lesion detection methods by using standard parameters. Increased True Detection Rate by 6.17% in parallel with decrease in Hammoud Distance by 2.3%, both compared to the best among alternative methods shows the effectiveness of the proposed scheme in detecting lesion borders of dermoscopy images.
Biomedical Image Processing / Medical Image Processing
Parisa Gifani; Hamid Behnam; Maryam Shojaee Fard
Volume 10, Issue 4 , January 2017, , Pages 303-313
Abstract
In this paper, we introduce a novel framework for illustrating the cardiac movements in echocardiogarphic images by utilizing temporal information and sparse representation. For the proposed method, we first derived temporal information by extracting intensity variation time curves (IVTC) assessed for ...
Read More
In this paper, we introduce a novel framework for illustrating the cardiac movements in echocardiogarphic images by utilizing temporal information and sparse representation. For the proposed method, we first derived temporal information by extracting intensity variation time curves (IVTC) assessed for each pixel. Then an over complete dictionary based on prior knowledge of the temporal signals and a set of pre-specified known functions was designed. The IVTCs can then be described as linear combinations of a few prototype atoms in the dictionary. We used the Bayesian Compressive Sensing (BCS) sparse recovery algorithm to find the sparse coefficients of the signals. By decomposing the IVTCs to different families and extracting proper features based on the sparse information, we attain the color coded images which illustrates the general movements of cardiac segments. The database consists of 21 echocardiography sequence of normal and abnormal volunteers in short axes and 4 chamber views. The results show the great achievement in global wall motion estimations.
Biomedical Image Processing / Medical Image Processing
maryam bagheri baghan; vahid azadzadeh; ali mohammad latif
Volume 10, Issue 2 , August 2016, , Pages 137-148
Abstract
It is a common approach to diagnose a disease based on the tongue in Traditional Chinese Medicine. In this paper, a noninvasive imaging of tongue whose surface papilla change in diabetics is used to detect the disease. The required images have been provided by Parsian specialized clinic of Mashhad. In ...
Read More
It is a common approach to diagnose a disease based on the tongue in Traditional Chinese Medicine. In this paper, a noninvasive imaging of tongue whose surface papilla change in diabetics is used to detect the disease. The required images have been provided by Parsian specialized clinic of Mashhad. In the sampling procedure, the diabetics, healthy individuals and those suspected of diabetes with both sexes and different age groups were considered. After imaging, tongue region was segmented based on two active contour models; then extended local binary patterns features, statistical features of the tongue texture, Color Moments in different color spaces were extracted from the segmented region. After feature extraction, diabetics, healthy and suspected of diabetes were detected using extreme learning machine classification. The proposed method obtained a precision of 97.7% for the current database. Experimental results show the efficiency and responding time of the proposed method compared to other noninvasive methods.
Biomedical Image Processing / Medical Image Processing
Seyedeh Zahra Islami Rad; Reza Gholipour Peyvandi
Volume 10, Issue 1 , May 2016, , Pages 49-57
Abstract
Positron emission tomography (PET) system is used in order to diagnose physiology changes in the body. Thus, the goal of the PET studies is to obtain a good quality and detailed image of organs by the PET scanner. The PET system performance and output image quality depend on the parameters including ...
Read More
Positron emission tomography (PET) system is used in order to diagnose physiology changes in the body. Thus, the goal of the PET studies is to obtain a good quality and detailed image of organs by the PET scanner. The PET system performance and output image quality depend on the parameters including spatial resolution, scatter fraction, sensitivity, RMS contrast and SNR which the system was evaluated based on them. In this paper, system features and tomography method for the IRI-microPET system are considered, firstly. Then, image reconstruction algorithms (MLEM, SART, and FBP) were performed on sinogram and the performance and the acquired images quality were evaluated. The radial and tangential resolutions of 1.81 mm and 1.90 mm for 18F at the center of FOV were measured. The scatter fraction of 7.1% for the mouse phantom and the sensitivity of 1.74% in 4 ns timing window was measured. Finally, images quality was compared by RMS contrast and SNR factors, which MLEM algorithm has superiorityin comparison with the other reconstructed algorithms.The acquired results from IRI-MicroPET system were compared with available commercial animal PET scanner which the results show the good agreement between data.
Biomedical Image Processing / Medical Image Processing
Fariborz Mahmoudi; Faraein Aeini
Volume 9, Issue 2 , July 2015, , Pages 113-131
Abstract
Due to teeth robustness, uniqueness and availability of medical records, today a new branch of research for human identification is ongoing based on dental radiograph images. This method of identification has particular importance especially in events such as wars, fires, tsunamis and other similar events ...
Read More
Due to teeth robustness, uniqueness and availability of medical records, today a new branch of research for human identification is ongoing based on dental radiograph images. This method of identification has particular importance especially in events such as wars, fires, tsunamis and other similar events with other biometrics features heavily damaged. In this article also a framework for human identification based on dental characteristics is proposed. The proposed framework consists of two stages: the first stage is teeth classification and numbering, and the second stage is teeth recognition. In this study, a new feature has been proposed for each of these two stages: Crown mesiodistal neck and anatomic crown length for the first and weighted sampling of teeth contours for the second. The proposed method is capable to solve principally and automatically problems such as diagnosis of posterior teeth, posterior teeth classification, diagnosis of number and kind of pulled teeth, which are overlooked or have been left with the simple premises in previous works. To evaluate the proposed method, experiments on a set of bitewings, periapical and panoramic images are done. The practical results show an improvement of 8% in accuracy of classification and numbering, and also 27% improvement in accuracy of teeth recognition, in comparison with the preceding works
Biomedical Image Processing / Medical Image Processing
Amir Ehsan Lashkari; Fatemeh Pak; Mohammad Firouzmand
Volume 9, Issue 1 , April 2015, , Pages 71-84
Abstract
Breast cancer is the most common type of cancer among women. The important key to treat the breast cancer is early detection of it because according to many pathological studies more 80% of all abnormalities are still benign at primary stages; so in recent years, many studies and extensive research done ...
Read More
Breast cancer is the most common type of cancer among women. The important key to treat the breast cancer is early detection of it because according to many pathological studies more 80% of all abnormalities are still benign at primary stages; so in recent years, many studies and extensive research done to early detection of breast cancer with higher precision and accuracy. Infra-red breast thermography is an imaging technique based on recording temperature distribution patterns of breast tissue. Compared with breast mammography technique, thermography is more suitable technique because it is noninvasive, non-contact, passive and free ionizing radiation. In this paper, a full automatic high accuracy technique for classification of suspicious areas in thermogram images with the aim of assisting physicians in early detection of breast cancer has been presented. Proposed algorithm consists of four main steps: pre-processing & segmentation, feature extraction, feature selection and classification. At the first step, using full automatic operation, region of interest (ROI) determined and the quality of image improved. Using thresholding and edge detection techniques, both right and left breasts separated from each other. Then relative suspected areas become segmented and image matrix normalized due to the uniqueness of each person's body temperature. At feature extraction stage, 23 features, including statistical, morphological, frequency domain, histogram and Gray Level Co-occurrence Matrix (GLCM) based features are extracted from segmented right and left breast obtained from step 1. To achieve the best features, feature selection methods such as mRMR, SFS, SBS, SFFS, SFBS and GA have been used at step 3. Finally to classify and TH labeling procedures, different classifiers such as AdaBoost, SVM, kNN, NB and PNN are assessed to find the best suitable one. The results obtained on native database showed the best and significant performance of the proposed algorithm in comprise to the similar studies. According to experimental results, mRMR combined with AdaBoost with the maximum accuracy of 92%, and SFFS combined with AdaBoost with a maximum accuracy of 88%, are the best combination of feature selection and classifier for evaluation of the right and left breast images respectively.
Biomedical Image Processing / Medical Image Processing
Somayeh Maleki Balajoo; Davoud Asemani; Hamid Soltanian-Zadeh
Volume 9, Issue 1 , April 2015, , Pages 99-111
Abstract
Although the cognitive deficits due to age-related brain differences have been largely analyzed, the altered connectivity of task related functional networks in aging requires more studies. As the brain of old adults experience some alterations in task performance during cognitive challenges, the related ...
Read More
Although the cognitive deficits due to age-related brain differences have been largely analyzed, the altered connectivity of task related functional networks in aging requires more studies. As the brain of old adults experience some alterations in task performance during cognitive challenges, the related effects on connectivity of functional networks are here evaluated using event-related functional Magnetic Resonance Imaging (fMRI). The fMRI data have been acquired for simple visual and motor tasks. For each subject, several Functional Connectivity (FC) networks including, motor, visual and the default mode networks are firstly calculated using a conventional voxel-wise correlation analysis with predefined region of interest. Then, the strength of functional connectivity is assessed and compared for different age groups. The current study has evaluated three hypotheses on FC of aging brain: the frontal regions involved with motor network try to compensate for declines in the posterior regions, default-mode network is less suppressed and, the posterior regions involved with visual network exhibit less connectivity. The first two hypotheses are accepted by analysis results but visual network behaves differently. Also, results show that the task related functional connectivity is considerably altered in old adults compared to young adults. Old adults demonstrate higher connectivity strength on average with a slightly smaller variance than young adults.