Accurate classification of images is essential for the analysis of mammograms in computer aided diagnosis of breast cancer. We propose a new approach to classify mammogram images based on fractal features. Given a mammogram image, we first eliminate all the artifacts and extract the salient features such as Fractal Dimension (FD) and Fractal Signature (FS). These features provide good descriptive values of the region. Second, a trainable multilayer feed forward neural network has been designed for the classification purposes and we compared the classification test results with K-Means. The result reveals that the proposed approach can classify with a good performance rate of 98%.

In this study, we present a three-stage method for detecting abnormalities and classifying electrocardiogram (ECG) beats using a k-nearest neighbor (k-NN) classifier and Gaussian mixture model (GMM). In the first stage, a signal filtering method is used to remove the ECG beat baseline wander. In the second stage, features are extracted based on Higuchi's fractal dimension (HFD) and statistical features. In the third stage, k-NN and GMM are used as classifiers to classify arrhythmia and ischemia. A total of 30,000 ECG segments obtained from the MIT-BIH Arrhythmia and European ST-T Ischemia databases were used to quantify this approach. 60% of the beats were used for training the classifier and the remaining 40%, for validating it. An overall accuracy of 99% and 98.24% was obtained for k-NN and GMM, respectively. This result is significantly better than that of currently used state-of-the-art classification approaches for arrhythmia and ischemia.

No abstract

No abstract

No abstract