P2791Method for detection of blood clots in ventricular assist devices

Le résumé fourni par la source

Abstract Background The technical development of modern ventricular assist devices (VAD) shows continued improvement in patient outcome. VADs continue to gain increased use not only as bridge to transplant but also as a destination therapy. However, VAD therapy is associated with complications, among which thromboembolic complications are common. Better management for early detection is paramount to improve patient outcome. Purpose The aims of this study were to determine if the VADs under consideration may retain clots, and to assess the feasibility of early clot detection using a non-invasive technique. Methods Two different VADs, the HeartMate II (HMII) and HeartMate 3 (HM3) were tested in a mock loop. Clot analogs (2 cm yarn fragments) were introduced into the loop upstream to the pump. Sound and vibration of the pump was measured using two hydrophones and at the inflow and outflow of the VAD (frequency range: 0.1Hz to 180kHz). The acoustic/vibration data was analyzed concerning the spectral content. Results The flow at the outflow section contained a low frequency swirling component. This appeared to be the basic flow feature. However, in the presence of clot analogs, the frequency signal was amplified. For the HMII this occurred below 10Hz. Concerning the HM3 this signal was more difficult to analyze due to HM3's non-constant rotational speed. However, by using wavelet filtering, it was possible to detect the low frequency signal in the presence of clot analogs. It was also observed that clot analogs tended to accumulate at certain regions in the respective VAD. In the HM3, clot fragments remained in a stagnation region at the bottom pump. This experimental finding contradicts published washout results based on a passive scalar model. Conclusions These mock loop results indicated the feasibility of early detection of blood clots in VADs by analysis of the sound/vibration induced by the clot. The signal “finger-print” needed for detection may also be possible to recover under non-continuous operational rotational speed as in the HM 3.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.