An Automatic Emboli Detection from Artifact using Fuzzy Multiagent System

Ayyoob Jafari and Bita Minaee

Keywords

Embolic Signal, Artifact, Fuzzy Multiagent, Genetic Algorithm

Abstract

Stroke is the second leading cause of death in the world today. Discrimination of embolic signals from artifacts makes it possible to detect patients with high risk of stroke, so that appropriate preventive measures are taken. Emboli are particles larger than normal cells in blood. Embolic signals are transient and are of short duration caused by presence of emboli. Usually Transcranial Doppler ultrasound (TCD) is used by specialists to detect emboli in blood. However, this manual process is painstakingly time consuming and is further complicated by presence of artifacts. In this research, we proposed a fuzzy multi-agent system based on suitable features such as AR, wavelet and Cepstral coefficients to improve detection of embolic signal from artifact. Results show the significant improvement in emboli detection, with 98% specificity and 97% accuracy.

Important Links:



Go Back