Title: Cardiovascular Health Informatics: Predicting the Risk of Stroke Based on Ultrasound Image Analysis of the Atherosclerotic Carotid Plaque
Abstract: Cardiovascular (CV) disease is one of the most common causes of death worldwide and represents a major financial burden for national economies. Effective prediction and prevention of CV disease particularly that, which resulted from high-risk asymptomatic atherosclerosis, has now become a top priority. The goal of this lecture will be to give a review of noninvasive ultrasound image processing methods that are used to facilitate the intelligent analysis of carotid plaque morphology for predicting stroke risk. The lecture will begin with a review of clinical methods for visual classification that have led to standardized methods for image acquisition. Then methods for ultrasound imaging atherosclerotic plaque denoising, and image segmentation will be described, followed by an overview of the several multi-scale texture-feature extraction algorithms and classification methods investigated. Then risk modeling based on clinical and ultrasonic plaque texture features that enable the assessment of the risk of stroke will be described.
2. Dr. Dimitris Tasoulis, Winton Capital Management, United Kingdom
Title: Data Clustering: From the Past to the Future
Abstract: Categorising entities into distinct and homogeneous groups has been an active research area since ancient Greece. In recent years, tasks like market basket analysis, image recognition and recommendation systems have provided a constant driving force for the development of data clustering as a scientific principle. This talk presents a top down view of the field, emphasising on the characteristics of each different application area. Starting from the early flat data tables and moving into the most recent streaming data format, it is shown how different data morphologies have affected the whole principle. The talk concludes with a discussion of the challenges that are yet to come.