An energy detector based on negative energy fluctuations coupled with sub-band decomposition as a robust micro-embolus detection system

摘要

Robust detection of the smallest circulating cerebral micro-emboli is an efficient way of preventing Cerebrovascular Accidents. Transcranial Doppler ultrasound is widely considered as the most convenient system for the detection of micro-emboli. Commercially realized standard detection is achieved through the whole Doppler energy spectrum where constant empirical thresholds are implemented. In this study, we propose a new type of micro-embolic detectors composed of N detectors associated to N Doppler frequency sub-bands. Moreover, we couple this sub-band decomposition with an original detection method. The N sub-band energy detectors are built from negative fluctuations of energy. From a training phase, we find that 4 sub-band decomposition allows obtaining the best results considering both the detection rate and the false alarm rate. The new sub-band detection coupled with energy detectors based from negative energy fluctuations allow increasing the detection rate up to 93% and the micro-embolus relative intensity increase from up to 23 dB. Moreover, the false alarm rate is reduced to around only 6%.