Zeljko Jovanovic

A comparison of methods for three-class mammograms classification

BACKGROUND Mammography is considered the gold standard for early breast cancer detection but it is very difficult to interpret mammograms for many reason. Computer aided diagnosis (CAD) is an important development that may help to improve the performance in breast cancer detection. OBJECTIVE We present a CAD system based on feature extraction techniques for detecting abnormal patterns in digital mammograms. METHODS Computed features based on gray-level co-occurrence matrices (GLCM) are used to evaluate the effectiveness of textural information possessed by mass regions. A total of 20 texture features are extracted from each mammogram. The ability of feature set in differentiating normal, benign and malign tissue is investigated using a Support Vector Machine (SVM) classifier, Naive Bayes classifier and K-Nearest Neighbor (k-NN) classifier. The efficiency of classification is provided using cross-validation technique. Support Vector Machine was originally designed for binary classification. We constructed a three-class SVM classifier by combining two binary classifiers and then compared his performance with classifiers intended for multi-class classification. To evaluate the classification performance, confusion matrix and Receiver Operating Characteristic (ROC) analysis were performed. RESULTS Obtained results indicate that SVM classification results are better than the k-NN and Naive Bayes classification results, with accuracy ratio of 65% according to 51.6% and 38.1%, respectively.The unbalanced classification that occurs in all three classification tests is reason for unsatisfactory accuracy. CONCLUSIONS Obtained experimental results indicate that the proposed three-class SVM classifier is more suitable for practical use than the other two methods.

Realization of Web GIS platform using open source technologies

Data which beside their own information contains support in the form of geographic information systems have much greater usability. Location of some object can be useful in a large number of uses of that information. This paper described open source platform as a Web application that provides an identification and permanent recording of desired location. Geographical support is provided through the Google Maps geographic information system. This paper describes the used platforms and the complete functionality of the created Web application.

Remotely analyze spine angle in rehabilitation after spine surgery using acceleration and gyro sensors

The purpose of this study was to determine range of motion values of lumbar spine in rehabilitation procedure after surgery, using wearable wireless sensors. In this paper, we present a method for determining the mobility of the spinal column using a network of sensors. The sensors consist of accelerometers and gyroscopes, and mutual communication is accomplished using a I2C bus. The main sensor node collects data from all the sensors and sends them to a computer using Bluetooth communication. The collected data is then filtered and converted to the values of the angles that are of interest to quantify the movement.

Android application for observing data streams from built-in sensors using RxJava

In modern applications it is necessary to process large amount of data from different sources in real time. These requirements can be met by data stream management system (DSMS) which provide reactive programming functionalities. Only some of them are lightweight and can be used in smart phones application. In this paper Android application with ReactiveX library is developed to detect uncomfortable driving events.

Building Web service for providing data from mBed based NCAP

Information which one transducer disposes may be useful for a large number of users from different software platforms. One of the ways that can offer informations to the Internet users is through Web services. For better interoperability, development of services in respect to standards is recommended. This paper aims to explain the process of creating a Web service that provides methods defined by the 1451.0 standard collecting data from (Network Capable Application Processor) NCAP developed on mBed development platform.

Test-Driven Development (TDD) in development of measure application and transducers services

Transducer services API provides the interface between the measure application and function defined by standrad IEEE 1451.0. This paper presents use Test-Driven Development (TDD) in implementation of methods of transducer services and Web application that access this services over the network. Characteristics of TDD and its benefits were presented and way of realization for one method was shown using Java and JUnit framework for run tests.

SunSpot in sevice-oriented applications

SunSpots — Java programmable devices with several built-in sensors allow the measurement and transmission of measured values between the Spots as the nodes in the network. In NetBeans development environment support to Sunspot desktop projects was developed. The theme of this paper is to expand the functionality of Sunspot desktop projects to support working with Web Services, which is not currently provided through NetBeans development environment and demonstration of SunSpots service-oriented inteoperability with Web Services.