Jafar M. H. Ali

Detection of spiculated masses in mammograms based on Fuzzy Image processing

This paper presents an efficient technique for the detection of spiculated massesin the digitized mammogram to assist the attending radiologist in making his decisions. The presented technique consists of two stages, enhancement of spiculation masses followed by the segmentation process. Fuzzy Histogram Hyperbolization (FHH) algorithm is first used to improve the quality of the digitized mammogram images. The Fuzzy C-Mean (FCM) algorithm is then applied to the preprocessed image to initialize the segmentation. Four measures of quantifying enhancement have been developed in this work. Each measure is based on the statistical information obtained from the labelled region of interest and a border area surrounding it. The methodology is based on the assumption that target and background areas are accurately specified. We have tested the algorithms on digitized mammograms obtained from the Digital Databases for Mammographic Image Analysis Society (MIAS).

Enhanced Rough Sets Rule Reduction Algorithm for Classification Digital Mammography

In this paper, we present an enhanced rough set approach for attribute reduction and generating classification rules from digital mammogram datasets. For this purpose, the presented approach is described in a hierarchical fashion. First, the preprocessing phase is adopted to enhance the contrast and edges of the mammogram images; moreover image processing segmentation algorithm is used to extract the region of interest. In the next phase, five texture features from the co-occurrence matrix are extracted and represented in attribute vector, and the reducts with minimal number of attributes are extracted. In the third phase, the decision rules within the generated reduct sets are generated. In the last phase, the classifier model was built and quadratic distances similarly function is used for matching process. To evaluate the validity of the rules based on the approximation quality of the attributes, we introduce a statistical test to evaluate the significance of the rules. The experimental results show that the classification algorithm performs well, reaching over 93% in accuracy with less number of rules compared with a well-known decision trees and neural network classifier models.

Digital mammogram segmentation algorithm using pulse coupled neural networks

This paper presents and develops an automated algorithm for segmenting speculated masses of the mammogram images based on pulse coupled neural networks (PCNN) in conjunction with fuzzy set theory. Mammogram image segmentation has proven to be a difficult task due to the low contrast between normal and malignant glandular tissues and the noise in such images that makes it very difficult to segment them. Therefore, the fuzzy histogram hyperbolization (FHH) algorithm is first used as a filter before the segmentation process. Then, the PCNN is applied to segment the images to arrive at the final result. To test the effectiveness of PCNNs on high quality images, a set of mammogram images was chosen. The experimental results show that the proposed algorithm performs well as compared to the fuzzy thresholds and fuzzy C-mean results.

Content-Based Image Classification and Retrieval: A Rule-based System Using Rough Sets Framework

Advances in data storage and image acquisition technologies have enabled the creation of large image datasets. Thus, it is necessary to develop appropriate information systems to efficiently manage these datasets. Image classification and retrieval is one of the most important services that must be supported by such systems. The most common approach used is content-based image retrieval (CBIR) systems. This paper presents a new application of rough sets to feature reduction, classification, and retrieval for image databases in the framework of content-based image retrieval systems. The suggested approach combines image texture features with color features to form a powerful discriminating feature vector for each image. Texture features are extracted, represented, and normalized in an attribute vector, followed by a generation of rough set dependency rules from the real value attribute vector. The rough set reduction technique is applied to find all reducts with the minimal subset of attributes associated with a class label for classification

Neural network model for database marketing in the new global economy

The discipline of marketing is going through significant changes. Such changes are necessitated by the globalization of markets aided by the facilitating communications and information processing technologies. In recent years, the discipline has also witnessed significant paradigm shifts such as relationship marketing and micromarketing. Information technology enabled marketers to accumulate large quantities of detailed information about their current and potential customers. However, to convert such vast databases effectively into useful management information, the traditional marketing research tools primarily based on parametric statistical methods are proving to be inadequate. In this context, in this paper it is argued that neural network models (NNMs) will prove to be robust methodological alternatives for marketers to practice database marketing effectively.

Rough set approach for classification of breast cancer mammogram images

This paper presents a study on classification of breast cancers in digital mammography images, using rough set theory in conjunction with statistical feature extraction techniques. First, we improve the contrast of the digitized mammograms by applying computer image processing techniques to enhance x-ray images and then subsequently extract features from suspicious regions characterizing the underlying texture of the breast regions. Feature extractions are derived from the gray-level co-occurrence matrix, then the features were normalized and the rough set dependency rules are generated directly from the real value attribute vector. These rules can then be passed to a classifier for discrimination for different regions of interest to test whether they are normal or abnormal. The experimental results show that the proposed algorithm performs well reaching over 98 % in accuracy.

An Empirical Examination of a Model of Computer Learning Performance

The growing sophistication of computer applications and the increasing diversity of end users have heightened the importance of and need for end user computer training. Therefore, understanding factors that influence learning performance in computer training continues to be an important issue for information systems research, education, and practice. This paper presents and empirically tests a model of learning performance in computer training. The model assesses the direct influence of computer self-efficacy, computer attitudes, and computer experience and the patterns of relationships among these variables in which they influence learning performance. The results indicate that computer self-efficacy and computer experience had direct and positive effects on learning performance. However, computer attitudes had indirect effect on learning performance through their direct effect on computer self-efficacy. Additionally, computer attitudes and computer experience had positive effects on computer self-efficacy. The results offer practical implications for selecting and preparing individuals for training and designing computer training content.

An Empirical Examination of Knowledge-Sharing Model in Information Systems Groups

Knowledge-sharing represents a key ingredient for group performance and success in work projects. Thus, understanding factors affecting knowledge-sharing in group work settings is critical for group effectiveness and success. Based on organisational behaviour and information systems (IS) literatures, the present study suggests that group efficacy and group cohesion will have direct effects on perceived loafing. In turn, perceived loafing, group efficacy, and group cohesion are posited to have direct effects on a members behavioural intention to share knowledge with other group members. The results revealed that group efficacy and group cohesion had negative effects on perceived loafing which, in turn, demonstrated a negative effect on behavioural intention to share knowledge. Group efficacy and group cohesion demonstrated non-significant direct effects on behavioural intention. These results provide valuable implications for research and practice.

Micro-Market Segmentation Using a Neural Network Model Approach

Abstract Market segmentation is one of the fundamental conceptual frameworks of the modern marketing theory and practice. Over time, several methods of market segmentation were proposed in the rich literature on the subject. In the past several years, a cluster analysis method was used to segment markets on various consumer behavioral bases. However, the more recent paradigm shift in marketing in favor of relationship marketing, growing importance of global marketing in diverse countries and focus on micro-market segments calls for better methodological tools for micro-market segmentation than the traditional parametric multivariate approaches with their methodological and practical limitations. In this context, in this paper it is proposed that a neural network model can be a viable alternative for micro-market segmentation. The neural network model applicability to micro-market segmentation was demonstrated with an empirical example.