15 research outputs found
Classification of juvenile myoclonic epilepsy data acquired through scanning electromyography with machine learning algorithms
No full textOsman, Onur (Arel Author), Özekes, Serhat (Arel Author)In this paper, classification of Juvenile Myoclonic Epilepsy (JME) patients and healthy volunteers included into Normal Control (NC) groups was established using Feed-Forward Neural Networks (NN), Support Vector Machines (SVM), Decision Trees (DT), and Na < ve Bayes (NB) methods by utilizing the data obtained through the scanning EMG method used in a clinical study. An experimental setup was built for this purpose. 105 motor units were measured. 44 of them belonged to JME group consisting of 9 patients and 61 of them belonged to NC group comprising ten healthy volunteers. k-fold cross validation was applied to train and test the models. ROC curves were drawn for k values of 4, 6, 8 and 10. 100% of detection sensitivity was obtained for DT, NN, and NB classification methods. The lowest FP number, which was obtained by NN, was 5
Automatic lung nodule detection using template matching
No full textWe have developed a computer-aided detection system for detecting lung nodules, which generally appear as circular areas of high opacity on serial-section CT images. Our method detected the regions of interest (ROIs) using the density values of pixels in CT images and scanning the pixels in 8 directions by using various thresholds. Then to reduce the number of ROIs the amounts of change in their locations based on the upper and the lower slices were examined, and finally a nodule template based algorithm was employed to categorize the ROIs according to their morphologies. To test the system's efficiency, we applied it to 276 normal and abnormal CT images of 12 patients with 153 nodules. The experimental results showed that using three templates with diameters 8, 14 and 20 pixels, the system achieved 91%, 94% and 95% sensitivities with 0.7, 0.98 and 1.17 false positives per image respectively
A Novel LSB Steganography Technique Using Image Segmentation
No full textSteganography is a process to hide data inside a cover file mostly used in media files like image, video, and audio files. Least significant bit (LSB) steganography is a technique where the least significant bits of pixels are used for information hiding. The purpose of using only those bits is to minimize the visual impact of the hidden data on the image file. LSB technique of steganography is one of the most popular forms of steganography available today. As a result, various steganalysis techniques are developed for this steganography technique. One of them is the visual analysis of pixels through pixel modifications to expose hidden data in a visual manner. The proposed method achieves resistance to this attack using an image segmentation model and extracting the most texture-complex areas of an image and hiding information in these specific areas as pseudo-randomized least significant bit replacements. As the outcome of the study, an alternative approach to LSB steganography that results competitively with existing methods is provided.
Improving cross-subject classification performance of motor imagery signals: a data augmentation-focused deep learning framework
No full textMotor imagery brain-computer interfaces (MI-BCIs) have gained a lot of attention in recent years thanks to their potential to enhance rehabilitation and control of prosthetic devices for individuals with motor disabilities. However, accurate classification of motor imagery signals remains a challenging task due to the high inter-subject variability and non-stationarity in the electroencephalogram (EEG) data. In the context of MI-BCIs, with limited data availability, the acquisition of EEG data can be difficult. In this study, several data augmentation techniques have been compared with the proposed data augmentation technique adaptive cross-subject segment replacement (ACSSR). This technique, in conjunction with the proposed deep learning framework, allows for a combination of similar subject pairs to take advantage of one another and boost the classification performance of MI-BCIs. The proposed framework features a multi-domain feature extractor based on common spatial patterns with a sliding window and a parallel two-branch convolutional neural network. The performance of the proposed methodology has been evaluated on the multi-class BCI Competition IV Dataset 2a through repeated 10-fold cross-validation. Experimental results indicated that the implementation of the ACSSR method (80.47%) in the proposed framework has led to a considerable improvement in the classification performance compared to the classification without data augmentation (77.63%), and other fundamental data augmentation techniques used in the literature. The study contributes to the advancements for the development of effective MI-BCIs by showcasing the ability of the ACSSR method to address the challenges in motor imagery signal classification tasks
Predicting Dropout Student: An Application of Data Mining Methods in an Online Education Program
No full textAbstract
This study examined the prediction of dropouts through data mining approaches in an online program. The subject of the study was selected from a total of 189 students who registered to the online Information Technologies Certificate Program in 2007-2009. The data was collected through online questionnaires (Demographic Survey, Online Technologies Self-Efficacy Scale, Readiness for Online Learning Questionnaire, Locus of Control Scale, and Prior Knowledge Questionnaire). The collected data included 10 variables, which were gender, age, educational level, previous online experience, occupation, self efficacy, readiness, prior knowledge, locus of control, and the dropout status as the class label (dropout/not). In order to classify dropout students, four data mining approaches were applied based on k-Nearest Neighbour (k-NN), Decision Tree (DT), Naive Bayes (NB) and Neural Network (NN). These methods were trained and tested using 10-fold cross validation. The detection sensitivities of 3-NN, DT, NN and NB classifiers were 87%, 79.7%, 76.8% and 73.9% respectively. Also, using Genetic Algorithm (GA) based feature selection method, online technologies self-efficacy, online learning readiness, and previous online experience were found as the most important factors in predicting the dropouts.</jats:p
Lung nodule diagnosis using 3D template matching
No full textIn this paper, to utilize the third dimension of Computed Tomography, regions of interest (ROI) slices were combined to form 3D ROI image and a 3D template was determined to find the structures with similar properties of nodules. Convolution of 3D ROI image with the proposed template strengthens the shapes similar to the template and weakens the other ones. False-positive (FP) per nodule and per slice versus diagnosis sensitivity were obtained. The Computer Aided Diagnosis system achieved 100% sensitivity with 0.83 FP per nodule and 0,46 FP per slice, when the nodule thickness was greater than or equal to 5.625 mm. (c) 2006 Elsevier Ltd. All rights reserved
ON EIGENVALUES OF THE LAPLACIAN AND CURVATURE OF RIEMANNIAN MANIFOLDS
No full textLet (M('n),g) be a compact connected orientable Riemannian manifold of dimension n,g the fundamental tensor. For 0 (LESSTHEQ) p (LESSTHEQ) n let (DELTA) be the Laplace-Beltrami operator on exterior p-form on M. The set.(DIAGRAM, TABLE OR GRAPHIC OMITTED...PLEASE SEE DAI).of eigenvalues of (DELTA) is the spectrum of M. The question of how the spectrum of M determines the structure of M has been studied by various authors. It is proved that the asymptotic expansion (the heat expansion).(DIAGRAM, TABLE OR GRAPHIC OMITTED...PLEASE SEE DAI).holds, where a(,i,p) are Riemannian invariants. The coefficients a(,0,0), a(,1,0), a(,2,0) have been computed by M. Berger and by McKean and Singer. a(,1,p), a(,2,p) have been computed by V. K. Patodi, and a(,3,0) by T. Sakai.(,).In this dissertation, for various special kinds of manifolds, we write out these coefficients and the coefficients b(,i)(i=1,2,3) of.(DIAGRAM, TABLE OR GRAPHIC OMITTED...PLEASE SEE DAI).the power series expansion (in normal coordinates) for the volume of a small geodesic ball with center m and radius r computed by A. Gray and L. Vanhecke. By forming tables for these coefficients we find several relations among them for some of the manifolds. We then compute the coefficients a(,3,p) and a(,4,p) for 2-dimensional compact Riemannian manifolds.Source: Dissertation Abstracts International, Volume: 42-06, Section: B, page: 2404.Ph.D. American University 1981.Englis
Psychometric Properties of the Future Time Perspective Scale for the Turkish Population: Age Differences in Predictors of Time Perspective
No full textThe purpose of this study was to determine the psychometric properties of the Turkish version of the Future Time Perspective Scale (FTPS-T) and examine age-group differences in the predictors of respondents’ future time perspective. Data were collected from a sample of 202 young adults (aged 18–28 years) and 127 community-dwelling older adults (aged 60–86 years). The internal consistency and test–retest methods were employed to assess the reliability of the FTPS-T, and the FTPS-T’s validity was assessed using construct- and criterion-related validity. The reliability and validity analyses demonstrated that the FTPS-T had satisfactory psychometric properties. Multiple regression analyses revealed that the strongest predictor of future time perspective in young adults was subjective psychological health, whereas chronological and subjective (i.e., physical) ages were stronger predictors among older adults. These findings indicate that subjective variables shape the perceptions of a lifetime, and the results are discussed in the context of socioemotional selectivity theory. © The Author(s) 2019
Nodule detection in a lung region that's segmented with using genetic cellular neural networks and 3D template matching with fuzzy rule based thresholding
No full textObjective: The purpose of this study was to develop a new method for automated lung nodule detection in serial section CT images with using the characteristics of the 3D appearance of the nodules that distinguish themselves from the vessels
