Water Cycle Bat Algorithm and Dictionary-Based Deformable Model for Lung Tumor Segmentation

Among the different types of cancers, lung cancer is one of the widespread diseases which causes the highest number of deaths every year. The early detection of lung cancer is very essential for increasing the survival rate in patients. Although computed tomography (CT) is the preferred choice for lungs imaging, sometimes CT images may produce less tumor visibility regions and unconstructive rates in tumor portions. Hence, the development of an efficient segmentation technique is necessary. In this paper, water cycle bat algorithm- (WCBA-) based deformable model approach is proposed for lung tumor segmentation. In the preprocessing stage, a median filter is used to remove the noise from the input image and to segment the lung lobe regions, and Bayesian fuzzy clustering is applied. In the proposed method, deformable model is modified by the dictionary-based algorithm to segment the lung tumor accurately. In the dictionary-based algorithm, the update equation is modified by the proposed WCBA and is designed by integrating water cycle algorithm (WCA) and bat algorithm (BA).

A muscle synergies-based movements detection approach for recognition of the wrist movements

Myoelectric signals are regarded as the control signal for prosthetic limbs. But, the main research challenge is reliable and repeatable movement detection using electromyography. In this study, the analysis of the muscle synergy pattern has been considered as a key idea to cope with this main challenge. The main objective of this research was to provide an analytical tool to recognize six wrist movements through electromyography (EMG) based on analysis of the muscle synergy patterns. In order to design such a system‚ the synergy patterns of the wrist muscles have been extracted and utilized to identify wrist movements. Also, different decision fusion algorithms were used to increase the reliability of the synergy pattern classification. The classification performance was evaluated while no data subject was enrolled. In terms of the achieved performance, using a multi-layer perceptron (MLP) neural network as the fusion algorithm turned out to be the best combination. The classification average accuracy, obtained in an offline manner, was about 99.78 ± 0.45%. While the classification average cross-validation accuracy, obtained in an offline manner, using Bayesian fusion, and Bayesian fuzzy clustering (BFC) fusion algorithm were 99.33 ± 0.80% and 96.43 ± 1.08%, respectively.

An Approach for Image Search and Retrieval by Cluster-Based Indexing of Binary MKSIFT Codes

Image search is an information retrieval approach that gained remarkable attention in the areas like multimedia and computer vision. The first work presented a cross-indexing-based approach for image retrieval using multiple kernel scale-invariant feature transform (MKSIFT), where the key point descriptor was calculated using two kernel functions. Previous work had complexity issues while dealing with the large databases, and hence, to avoid this, cluster-based indexing of binary MKSIFT codes is presented. The proposed cluster-based indexing scheme uses the MKSIFT feature extraction and the Particle Swarm-Fractional Bacterial foraging optimization algorithm for extracting the useful features from the images. Also, the Bayesian fuzzy clustering scheme is employed for grouping the images in the database into several clusters. The search index is constructed for the user query, and Bhattacharya distance between the cluster centroids and the search index is calculated to identify the optimal cluster. Then, finally, the images present in the optimal centroid are retrieved. The experimentation of the proposed cluster-based indexing scheme is analysed for the various query from the users. From the analysis, it is evident that the proposed cluster-based indexing scheme has achieved improved performance with the mean values of 0.9656, 0.9489 and 0.9049, for precision, recall and F measure, respectively.

WS-BD-Based Two-Level Match: Interesting Sequential Patterns and Bayesian Fuzzy Clustering for Predicting the Web Pages from Weblogs

The rapid increase in information and technology has led to the increased amount of web pages, which raises the complexity in sticking to relevant web pages, and the visitor suffers due to wastage of time resulting in lack of satisfaction. This paper proposes a web page prediction method using a weighed support and Bhattacharya distance-based (WS-BD) two-level match. The major aim of the proposed method is to attain customer satisfaction. Initially, interesting sequential patterns are obtained using the weighed support that filters the sequential patterns obtained using a PrefixSpan algorithm based on the frequency, duration and recurrence of the web pages. Interesting sequential patterns are clustered using the proposed dice similarity-based Bayesian fuzzy clustering, and the web page is predicted using the two-level match based on Bhattacharya distance. The experimentation is performed using the CTI and MSNBC data which proves the effectiveness of the proposed method. The proposed method shows 9.59, 21.22 and 10.17% improvement than the existing FCM-KNN in terms of precision, recall and F measure for the CTI dataset. Also, the proposed method shows 2.58, 22.17 and 7.83% improvement than the existing FCM-KNN in terms of precision, recall and F measure for the MSNBC dataset.