Turgay Ibrikci

Effective Diagnosis of Coronary Artery Disease Using The Rotation Forest Ensemble Method

Coronary Artery Disease is a common heart disease related to disorders effecting the heart and blood vessels. Since the disease is one of the leading causes of heart attacks and thus deaths, diagnosis of the disease in its early stages or in cases when patients do not show many of the symptoms yet has considerable importance. In the literature, studies based on computational methods have been proposed to diagnose the disease with readily available and easily collected patient data, and among these studies, the greatest accuracy reached is 89.01%. This paper presents a computational tool based on the Rotation Forest algorithm to effectively diagnose Coronary Artery Disease in order to support clinical decision-making processes. The proposed method utilizes Artificial Neural Networks with the Levenberg-Marquardt back propagation algorithm as base classifiers of the Rotation Forest ensemble method. In this scheme, 91.2% accuracy in diagnosing the disease is accomplished, which is, to the best of our knowledge, the best performance among the computational methods from the literature that use the same data. This paper also presents a comparison of the proposed method with some other classifiers in terms of diagnosis performance of Coronary Artery Disease.

PCA based clustering for brain tumor segmentation of T1w MRI images

BACKGROUND AND OBJECTIVE Medical images are huge collections of information that are difficult to store and process consuming extensive computing time. Therefore, the reduction techniques are commonly used as a data pre-processing step to make the image data less complex so that a high-dimensional data can be identified by an appropriate low-dimensional representation. PCA is one of the most popular multivariate methods for data reduction. This paper is focused on T1-weighted MRI images clustering for brain tumor segmentation with dimension reduction by different common Principle Component Analysis (PCA) algorithms. Our primary aim is to present a comparison between different variations of PCA algorithms on MRIs for two cluster methods. METHODS Five most common PCA algorithms; namely the conventional PCA, Probabilistic Principal Component Analysis (PPCA), Expectation Maximization Based Principal Component Analysis (EM-PCA), Generalize Hebbian Algorithm (GHA), and Adaptive Principal Component Extraction (APEX) were applied to reduce dimensionality in advance of two clustering algorithms, K-Means and Fuzzy C-Means. In the study, the T1-weighted MRI images of the human brain with brain tumor were used for clustering. In addition to the original size of 512 lines and 512 pixels per line, three more different sizes, 256 × 256, 128 × 128 and 64 × 64, were included in the study to examine their effect on the methods. RESULTS The obtained results were compared in terms of both the reconstruction errors and the Euclidean distance errors among the clustered images containing the same number of principle components. CONCLUSION According to the findings, the PPCA obtained the best results among all others. Furthermore, the EM-PCA and the PPCA assisted K-Means algorithm to accomplish the best clustering performance in the majority as well as achieving significant results with both clustering algorithms for all size of T1w MRI images.

Diagnosis of Several Diseases by Using Combined Kernels with Support Vector Machine

Machine learning techniques have gained increasing demand in biomedical research due to capability of extracting complex relationships and correlations among members of the large data sets. Thus, over the past few decades, scientists have been concerned about computer information technology to provide computational learning methods for solving the complex medical problems. Support Vector Machine is an efficient classifier that is widely applied to biomedical and other disciplines. In recent years, new opportunities have been developed on improving Support Vector Machines’ classification efficiency by combining with any other statistical and computational methods. This study proposes a new method of Support Vector Machines for influential classification using combined kernel functions. The classification performance of the developed method, which is a type of non-linear classifier, was compared to the standart Support Vector Machine method by applying on seven different datasets of medical diseases. The results show that the new method provides a significant improvement in terms of the probability excess.

Fuzzy C-Means Based DNA Motif Discovery

In this paper, we examined the problem of identifying motifs in DNA sequences. Transcription-binding sites, which are functionally significant subsequences, are considered as motifs. In order to reveal such DNA motifs, our method makes use of Fuzzy clustering of Position Weight Matrix. The Fuzzy C-Means (FCM) algorithm clearly predicted known motifs that existed in intergenic regions of GAL4, CBF1 and GCN4 DNA sequences. This paper also provides a comparison of FCM with some clustering methods such as Self-Organizing Map and K-Means. The results of the FCM algorithm is compared to the results of popular motif discovery tool Multiple Expectation Maximization for Motif Elicitation (MEME) as well. We conclude that soft-clustering-based machine learning methods such as FCM are useful to finding patterns in biological sequences.

Gravitational Fuzzy Clustering

Data clustering is an important part of cluster analysis. Numerous clustering algorithms based on various theories have been developed, and new algorithms continue to appear in the literature. In this paper, supposing that each cluster center is a gravity center and each data point has a constant mass, Newtons law of gravity is transformed from m/d2to 1/d2. According to adapted the law, we have proposed novel method called Gravitational Fuzzy clustering. The three main contributions of new algorithm can be summarized as: 1) it becomes more sophisticated technique by taking advantages of K-means, fuzzy C-means and subtractive clustering methods, 2) it removes the dependence on initial condition by taking account of the gravitation effect, 3) it improves the cluster centers by means of the gravity center of clusters. We illustrate the advantage of the resulting of gravitational approach with several examples.

Prediction of disorder with new computational tool: BVDEA

Recognizing that many intrinsically disordered regions in proteins play key roles in vital functions and also in some diseases, identification of the disordered regions has became a demanding process for structure prediction and functional characterization of proteins. Therefore, many studies have been motivated on accurate prediction of disorder. Mostly, machine learning techniques have been used for dealing with the prediction problem of disorder due to the capability of extracting the complex relationships and correlations hidden in large data sets. In this study, a novel method, named Border Vector Detection and Extended Adaptation (BVDEA) was developed for predicting disorder as an alternative accurate classifier. The classifier performs the predictions by using three types of structural features belonging to proteins. For attesting the performance of the method, three computational learning techniques and eleven specific tools were used for comparison. Training was executed based on the data by 5-fold cross validation. When compared with the two learning methods of LVQ and BVDA, the proposed method gives the best success on classification. The BVDEA also provides faster and more robust learning as compared to the others. The new method provides a significant contribution to predicting disorder and order regions of proteins.

Consensual classification of drug and nondrug compounds

A special consensual approach is discussed for separating a molecular group with a proven pharmacological activity from another molecular group without any activity. It is mainly a group decision to produce a consensus of multiple classification results obtained with a single classification algorithm. For this purpose, the constructed model has a preprocessing unit which consists of transformation of input patterns by random matrices and median filtering to generate independent errors for a single type of classifier and postprocessing for consensus. The neural network based consensus classifier operating with MOE descriptors was applied to a set of 641 chemical structures. The confirmed drugs were classified with an accuracy of 86.54% while nondrugs resulted in 82.67% accuracy.

Emphysema discrimination from raw HRCT images by convolutional neural networks

Emphysema is a chronic lung disease that causes breathlessness. HRCT is the reliable way of visual demonstration of emphysema in patients. The fact that dangerous and widespread nature of the disease require immediate attention of a doctor with a good degree of specialized anatomical knowledge. This necessitates the development of computer-based automatic identification system. This study aims to investigate the deep learning solution for discriminating emphysema subtypes by using raw pixels of input HRCT images of lung. Convolutional Neural Network (CNN) is used as the deep learning method for experiments carried out in the Caffe deep learning framework. As a result, promising percentage of accuracy is obtained besides low processing time.

A Bayesian Scoring Scheme based Particle Swarm Optimization algorithm to identify transcription factor binding sites

Identification of transcription factor binding sites is a vital task in contemporary biology, since it helps researchers to comprehend the regulatory mechanism of gene expression. Computational tools to perform this task have gained great attention since they are good alternatives to expensive and laborious biological experiments. In this paper, we propose a Particle Swarm Optimization based motif-finding method that utilizes a proven Bayesian Scoring Scheme as the fitness function. Since PSO is designed to work in multidimensional continuous domains, this paper presents required developments to adapt PSO for the motif finding application. Furthermore, this paper presents a benchmark of PSO variants with four separate population topologies, GBest, Bidirectional Ring, Random and Von Neumann. Simulations performed over synthetic and real data sets have shown that the proposed method is efficient and also superior to some well-known existing tools. Additionally, the Bidirectional Ring topology appears to be remarkable for the motif-finding application.

Assessment of clustering algorithms for unsupervised transcription factor binding site discovery

Identification of transcription factor binding sites is a key task to understand gene regulation mechanism to discover gene networks and functions. Clustering approach is proved to be useful when finding such patterns residing in promoter regions of co-regulated genes. Four clustering algorithms, Self-Organizing Map, K-Means, Fuzzy C-Means and Expectation-Maximization are studied in this paper to discover motifs in datasets extracted from Saccharomyces cerevisiae, Escherichia coli, Droshophila melanogaster and Homo sapiens DNA sequences. Required modifications to clustering algorithms in order to adapt them to motif finding task are presented through the paper. Then, their motif-finding performances are discussed carefully and evaluated against a popular motif-finding method, MEME.