Search

Summary Background: Stroke is a major complication of nonvalvular atrial fibrillation (AF). Despite its proven efficacy in stroke prevention, oral anticoagulation (OAC) is associated to a significant increase in bleeding complications. New techniques such as percutaneous left atrial appendage (LAA) closure were developed. Aim: To evaluate immediate, mid- and long-term outcomes after percutaneous LAA closure in Tunisian patients presenting with nonvalvular AF. Methods and Results: Nineteen patients with nonvalvular AF were prospectively enrolled for percutaneous LAA closure between February 2013 and June 2014. The Watchman device was used in all LAA closure procedures. Clinical and echocardiographic follow-up were carried-out at 1, 6, 12 months and six years in all patients. Mean age was 68.4 ± 7.5 years. Thirteen patients were female, 16 had hypertension, 12 had diabetes mellitus and seven had a history of stroke or transient ischemic attack (TIA). Average CHA2DS2VASc (Congestive Heart Failure/Left Ventricular Dysfunction, Hypertension, Age≥75, Diabetes mellitus, Stroke/Transient Ischemic Attack/Thromboembolic event, Vascular disease,Age≥65, Sex category) score was 4.2 ± 1.5 and HAS-BLED (Hypertension, Abnormal renal/liver function, Stroke, Bleeding tendency, Labile INR, Age≥65, Drugs) score was 3.5 ± 1. Ten patients had a history of severe bleeding. Procedural success was achieved in all patients. Pericardial effusion with tamponade was reported in one case. No post-procedural death was reported. Regular follow-up at 1, 6 and12 months then every year up to 6 years reported no stroke, no thromboembolic event, no Watchman device thrombosis and three cases of death caused by a respiratory problem and cancers. Conclusion According to this study, LAA closure with Watchman device was safe and effective in preventing stroke in patients with nonvalvular AF and contra indication to OAC.

IoT data collection networks have recently become one of the important research areas due to their fundamental role and wide application in many domains. The establishment of networks of objects is based essentially on the deployment of connected objects to process the collected data and transmit them to the various locations. Subsequently, a large number of nodes must be adequately deployed to achieve complete coverage. This manuscript introduces a distributed approach, which combines the Voronoi Diagram and the Genetic algorithm(VD-GA), to maximize the coverage of a region of interest. The Voronoi diagram is used to divide region into cells and generate initial solutions that present the positions of the deployed IoT objects. Then, a genetic algorithm is executed in parallel in several nodes to improve these positions. The developed VD-GA approach was evaluated on an experimental environment by prototyping on a real testbed utilizing M5StickC nodes equipped with ESP32 processor. The experiments show that the distributed approach provided better degree of coverage, RSSI, lifetime and number of neighboring objects than those given by the original algorithms in terms of the suggested distributed Genetic-Voronoi algorithm outperforms the centralized one in terms of speed of computing. .

International audience; IoT data collection networks, formerly known as wireless sensor networks, have become one of the most active areas of research in the field of information technology. The deployment of connected objects represents a fundamental phase on the establishment of IoT collection networks. The IoT deployment problem consists in positioning all the sensors constituting the network. This paper proposes an approach maximizing the coverage of a region of interest by the hybridization between the Voronoi diagram and the genetic algorithm. The first algorithm divides the field into cells and generates initial solutions (positions of deployed IoT objects). The latter algorithm is used to improve these positions in order to maximize the overall coverage of the region of interest. Obtained results reveal that the performance of the hybrid algorithm exceeds that of the original algorithms in terms of coverage degree, RSSI, lifetime and number of neighbor of objects.