فهرست مطالب

Journal of Medical Signals and Sensors
Volume:3 Issue: 2, Apr-Jun 2013

  • تاریخ انتشار: 1392/02/03
  • تعداد عناوین: 8
|
  • Golshan Taheri, Mehran Yazdi, Alireza Keshavarz, Haddad, Arash Rafie Borujeny Page 63
    The monitoring of epileptic seizures is mainly done by means of EEG (Electroencephalogram) monitoring. Although this method is accurate, it is not comfortable for the patient as the EEG-electrodes have to be attached to the scalp which hampers the patient’s movement. This makes long term home monitoring not feasible.In this paper the aim is to propose a seizure detection system based on accelerometry for the detection of epileptic seizure. The used sensors are wireless, which can improve quality of life for patients. In this system three 2D accelerometer sensors are positioned on right arm, left arm and left thigh of an epileptic patient. Datasets from three patients suffering from severe epilepsy are used in this paper for the development of an automatic detection algorithm.This monitoring system is based on Wireless Sensor Networks that can determine the location of the patient when a seizure is detected and sends an alarm to hospital staff or their relatives. Our wireless sensor nodes are MICAz Motes developed by Crossbow Technology. The proposed system can be used for patients living in a clinical environment or their home, where they do only their daily routines.The analysis of the recorded data is done by Artificial Neural Networks and K nearest-neighbor to recognize seizure movements from normal movements. The results show that the best algorithm for seizure detection is K Nearest Neighbor. We have shown that if at least 50 percent of signals consist of seizure samples, we can detect seizure accurately. Consequently, there is no need for training the algorithm for each new patient.
  • Mahdieh Ghasemi, Ali Mahloojifar Page 69
    Parkinson’s disease (PD) is a progressive neurological disorder characterized by tremor, rigidity, and slowness of movements. Specific changes associated with various pathological attacks in Parkinson’s Disease can be indicated in causality interactions of the brain Network from resting state fMRI data. In this paper, we aimed to reveal the network architecture of the directed influence brain network using multivariate Granger causality analysis and graph theory in patients with PD compared with control group. Functional magnetic resonance imaging (rs-fMRI) at rest from 10 PD patients and 10 controls were analyzed. Topological properties of the networks showed that, flow of information in PD is smaller than healthy. We found that there is a balanced local network in healthy include positive pair-wise cross connections between Caudate and Cerebellum and, reciprocal connections between motor cortex and caudate in left and right hemispheres. The results showed that this local network is disrupted in PD due to disturbance of the interactions in the motor networks. These findings suggested alteration of the functional management of the brain in the resting state, that affect the information organization from and to the other brain regions related to both primary dysfunctions and higher-level cognition impairments in PD. Furthermore, we showed that ROIs with high degree values could detect as betweenness centrality nodes. Our results demonstrate that properties of small-world connectivity could also recognize and quantify the characteristics of directed influence brain networks in Parkinson’s disease.
  • Dr Alireza Mehri Dehnavi, Mohammadreza Sehhati, Dr Hossein Rabbani Page 79
    Background
    Using primary tumor gene expression has been shown to have the ability of finding metastasis-driving gene markers for the prediction of breast cancer recurrence (BCR). However, there are some difficulties associated with the analysis of microarray data which led to poor predictive power and inconsistency of the previously introduced gene signatures.
    Methods
    In this study a hybrid method was proposed for identifying more predictive gene signatures from microarray datasets. Initially, the parameters of a Rough-Set (RS) theory based feature selection method were tuned to construct a customized gene extraction algorithm. Afterward, using the RS gene selection method the most informative genes were selected from six independent breast cancer datasets. Then, the combined set of these six signature sets, containing 114 genes, was evaluated for the prediction of BCR. Finally, a meta-signature, containing 18 genes, was selected from the combination of datasets and its prediction accuracy was compared with the combined signature.
    Results
    The results of 10-fold cross validation test, showed acceptable misclassification error rate (MCR) over 1338 cases of breast cancer patients. In comparison with a recent similar work our approach reached more than 5% reduction in MCR using a fewer number of genes for the prediction. The results also demonstrated 7% improvement in the average accuracy in six utilized datasets, using the combined set of 114 genes in comparison with the 18-genes meta-signature.
    Conclusions
    In this study, a more informative gene signature was selected for the prediction of BCR using a RS based gene extraction algorithm. To conclude, combining different signatures demonstrated more stable prediction over the independent datasets.
  • Mohammadreza Sehhati, Alireza Mehri Dehnavi, Hossein Rabbani, Shaghayegh Haghjoo Javanmard Page 87
    Background
    Numerous studies used microarray gene expression data to extract metastasis-driving gene signatures for the prediction of breast cancer relapse. However, the accuracy and generality of the previously introduced biomarkers are not acceptable for reliable usage in independent datasets. This inadequacy is attributed to ignoring gene interactions by simple feature selection methods, due to their computational burden.
    Materials And Methods
    In this study, an integrated approach with low computational cost was proposed for identifying a more predictive gene signature, for prediction of breast cancer recurrence. First, a small set of genes was primarily selected as signature by an appropriate filter feature selection (FFS) method. Then, a binary sub-class of protein–protein interaction (PPI) network was used to expand the primary set by adding adjacent proteins of each gene signature from the PPI-network. Subsequently, the support vector machine-based recursive feature elimination (SVMRFE) method was applied to the expression level of all the genes in the expanded set. Finally, the genes with the highest score by SVMRFE were selected as the new biomarkers.
    Results
    Accuracy of the final selected biomarkers was evaluated to classify four datasets on breast cancer patients, including 800 cases, into two cohorts of poor and good prognosis. The results of the five-fold cross validation test, using the support vector machine as a classifier, showed more than 13% improvement in the average accuracy, after modifying the primary selected signatures. Moreover, the method used in this study showed a lower computational cost compared to the other PPI-based methods.
    Conclusions
    The proposed method demonstrated more robust and accurate biomarkers using the PPI network, at a low computational cost. This approach could be used as a supplementary procedure in microarray studies after applying various gene selection methods.
  • Reza Azmi, Boshra Pishgoo, Narges Norozi, Samira Yeganeh Page 94
    Brain MR images tissue segmentation is one of the most important parts of the clinical diagnostic tools. Pixel classification methods have been frequently used in the image segmentation with two supervised and unsupervised approaches up to now. Supervised segmentation methods lead to high accuracy but they need a large amount of labeled data, which is hard, expensive and slow to obtain. Moreover, they cannot use unlabeled data to train classifiers. On the other hand, unsupervised segmentation methods have no prior knowledge and lead to low level of performance. However, semi-supervised learning (SSL) which uses a few labeled data together with a large amount of unlabeled data causes higher accuracy with less trouble. In this paper, we propose an ensemble semi-supervised framework for segmenting of brain MRIs tissues that it has been used results of several semi-supervised classifiers simultaneously. Selecting appropriate classifiers has an important role in the performance of this framework. Hence, in this paper we present two semi-supervised algorithms EFM and MCo_Training that are improved versions of semi-supervised methods EM and Co_Training and increase segmentation accuracy. Afterwards, we use these improved classifiers together with Graph-Based semi-supervised classifier as components of the ensemble framework. Experimental results show that performance of segmentation in this approach is higher than both supervised methods and the individual semi-supervised classifiers
  • Dr Keyvan Jabbari, Nazli Azarmahd, Dr Shadi Babazadeh, Dr Alireza Amooheidari Page 107
    Radiotherapy plays an essential role in the management of breast cancer. Three-dimensional conformal radiation therapy (3D-CRT) is applied based on 3D image information of anatomy of patients. In three-dimensional conformal radiation therapy (3D-CRT) for breast cancer one of the common techniques is tangential technique. In this project, various parameters of tangential and supraclavicular fields are optimized.This project has been done on CT images of 100 patients in Isfahan Milad Hospital. All patients have been simulated and all the important organs have been contoured by radiation oncologist. Two techniques in supraclavicular region are evaluated including: 1- A single field (AP Anterior Posterior) with a dose of 200cGy per fraction with 6MV energy. This is a common technique. 2- Two parallel opposed fields (Anterior Posterior-Posterior Anterior). The dose of AP was 150cGy with 6MV energy and PA 50cGy with 18MV. In the second part of the project, the tangential fields has been optimized with change of normalization point in five points: 1- Iso center (Confluence of rotation gantry axis and collimator axis) 2- Middle of thickest part of breast or middle of inter field distance (IFD) 3- Border between lung and chest wall 4- Physician’s choice 5- Between IFD and isocenter.Dose distributions have been compared for all patients in different methods of supraclavicular and tangential field. In parallel opposed fields average lung dose was 4% more than single field and the maximum received heart dose was 21.5% less than single field.The average dose of PTV (Planning Tumor Volume) in method 2 is 2 % more than method 1. In general AP-PA method because of a better coverage of PTV is suggested.In optimization of the tangential field all methods have similar coverage of PTV. Each method has spatial advantages and disadvantages. If it is important for the physician to reduce the dose received by the lung and heart, fifth method is suggested since in this method average and maximum received dose to heart and lung have been reduced few percent in comparison to other methods. If a better coverage of PTV is important for the physician second method can be an optimized method. In this method average and maximum received dose to PTV have been increased few percent in comparisons of physician’s choice method and three other methods.In optimizing of supraclavicular field AP-PA method due to better coverage of PTV is suggested. In optimizing of tangential all methods are similar. Each method has special advantages and disadvantages. The physicians can change the depth of the normalization point in the breast to get the desired average dose.
  • Azardokht Amirzadi, Reza Azmi Page 117
    Since mammography images are in low-contrast, applying enhancement techniques as a pre-processing step are wisely recommended in classification of the abnormal lesions into benign or malignant. A new kind of structural enhancement is proposed by morphological operator which introduces an optimal Gaussian Kernel primitive, the kernel parameters are optimized the use of Genetic Algorithm­­. We also take the advantageous of Optical Density (OD) images to promote the diagnosis rate. OD images are free from scanner type, and their values are the degree of blackness presented at the given point on the film and distinguish small differences. When the proposed enhancement method is applied on both the Gray Level (GL) images and their OD values respectively, morphological patterns get bolder on gray level images, therefore; Local Binary Patterns (LBP) are extracted from this kind of images. Applying the enhancement method on OD images causes to remove some background pixels. Those pixels that are more eligible to be mass are remained, and some statistical texture features are extracted. Support Vector Machine is used for both approaches, and the final decision is made by combining these two classifiers. The classification performance rate is evaluated by Az, under the receiver operating characteristic (ROC) curve. The designed method yields A­­­z = 0.9231 which demonstrates good results.
  • Milad Baradaran, Jaffar Fattahi Asl, Molood Baradaran, Mojtaba Karbalae, Hamid Reza Baradaran Page 127
    Reply to the letter sent by Prof. Viroj Wiwanitkit entitled “Radiofrequency Radiation and Human Ferritin”