Service Communautaire d'Information sur la Recherche et le Développement - CORDIS

Automatic embolus detection system

Identification of microemboli in the cerebral circulation using transcranial Doppler (TCD) ultrasound provides valuable clinical information, but currently micro-embolic signal (MES) detection and analysis are significantly limited as they rely mainly on costly off-line manual analysis by human experts. A knowledge-based system has been designed and implemented for automated identification and archival of MESs in real-time. Preliminary tests have been carried out to evaluate the system.

The system hardware consists of a PC and a powerful digital signal processing (DSP) board. The signal-processing engine of the system contains an FFT-based spectral analysis unit and a MES detection unit utilizing expert system reasoning theory. The system extracts information in time, frequency and neighbourhood domains from input Doppler signals. The information is then further processed using an expert system blackboard structure. For MES archival, segments of the raw Doppler signal corresponding to each detected event are saved into a data file, together with all relevant information. Preliminary clinical tests have been carried out using Doppler signals obtained from carotid endarterectomy patients and healthy volunteers. The results have demonstrated a sensitivity of 93.6% and a specificity of 99.2%.

Informations connexes

Reported by

University of Leicester
Department of Medical Physics, Leicester Royal Infirmary.
LE1 5WW Leicester
United Kingdom
See on map