Identification of echocardiographic "smoke" in a bench model with transcranial Doppler ultrasound.

BACKGROUND AND PURPOSE Spontaneous echo contrast in cardiac chamber has been indicated as a source of cerebral embolism. The nature of the echocardiographic smokelike signal is still not fully understood. This study was designed to regenerate spontaneous echo contrast and verify its thromboembolic characters in an in vitro model. METHODS Spontaneous echo contrast was reproduced in an expansion chamber under low flow conditions in a close circulation system. The spontaneous echo contrast was monitored and recorded with a 2-dimensional cardiosonography system and a transcranial Doppler device. Meanwhile, clinically commonly encountered embolic materials such as whole-blood clots, platelet aggregate-rich plasma, air bubbles, and 100-mL normal saline were injected into this monitored circuit. The differentiation of spontaneous echo contrast from emboli was performed by both visual observations of the echo images and offline Doppler signal intensity analysis. Average signal intensities produced by spontaneous contrast and injection of embolic materials and saline were compared. Furthermore, the effect of Doppler-detected flow velocity on generation of spontaneous contrast was also evaluated. RESULTS Spontaneous echo contrast was reproduced at low flow settings (90 to 120 mL/min) in this model. There was no significant difference in average signal intensity between the flow with spontaneous echo contrast and that without the echo (P=0.71). However, injection of embolic materials or normal saline did not generate smokelike image but caused much higher average signal intensity than the flow with spontaneous contrast (P<0.001). Injection of normal saline also increased average signal intensity. CONCLUSIONS Our results suggest that smokelike echo is a special echo phenomenon occurring at low flow situations and does not itself produce material capable of embolizing into the systemic circulation.