BACKGROUND: As outcomes for destination therapy continue to improve, many patients are requiring left ventricular assist device (LVAD) exchange due to end-of-life of their LVAD. Current techniques to identify and diagnose device end-of-life issues usually require invasive testing or off-site filter dust analysis. In this study we assess a non-invasive technique using acoustic signals generated from the HeartMate XVE LVAD to potentially identify impending device end-of-life issues. METHODS: Nine patients were prospectively followed after implantation of the HeartMate XVE LVAD as destination therapy between May 2004 and July 2006. Acoustic signals were collected using an aquatic hydrophone system interfaced with a data acquisition system and a standard laptop computer. Data were collected at pre-set intervals. All data/acoustic signals were prospectively interpreted by a blinded independent reviewer skilled at interpreting acoustic signals. Acoustic data suggesting possible device failure were then correlated with clinical findings and LVAD examination at the time of device removal. RESULTS: All patients survived long enough to develop signs of impending device end-of-life. Four of 9 (44%) patients developed inflow valve incompetence, 4 (44%) were identified as having significant bearing wear, and 1 (12%) had both. All acoustically identified device issues were confirmed by standard clinical examinations and testing (echocardiography, angiography, laboratory tests and filter dust analysis). The acoustic findings were subsequently confirmed at time of device exchange. All patients ultimately had their device successfully exchanged and have continued to live with their new apparatus. CONCLUSIONS: Acoustic signal monitoring can successfully identify HeartMate XVE device end-of-life. This new method provides a low-cost, reproducible, non-invasive technique that may be used to identify possible impending device failure.
BACKGROUND: As outcomes for destination therapy continue to improve, many patients are requiring left ventricular assist device (LVAD) exchange due to end-of-life of their LVAD. Current techniques to identify and diagnose device end-of-life issues usually require invasive testing or off-site filter dust analysis. In this study we assess a non-invasive technique using acoustic signals generated from the HeartMate XVE LVAD to potentially identify impending device end-of-life issues. METHODS: Nine patients were prospectively followed after implantation of the HeartMate XVE LVAD as destination therapy between May 2004 and July 2006. Acoustic signals were collected using an aquatic hydrophone system interfaced with a data acquisition system and a standard laptop computer. Data were collected at pre-set intervals. All data/acoustic signals were prospectively interpreted by a blinded independent reviewer skilled at interpreting acoustic signals. Acoustic data suggesting possible device failure were then correlated with clinical findings and LVAD examination at the time of device removal. RESULTS: All patients survived long enough to develop signs of impending device end-of-life. Four of 9 (44%) patients developed inflow valve incompetence, 4 (44%) were identified as having significant bearing wear, and 1 (12%) had both. All acoustically identified device issues were confirmed by standard clinical examinations and testing (echocardiography, angiography, laboratory tests and filter dust analysis). The acoustic findings were subsequently confirmed at time of device exchange. All patients ultimately had their device successfully exchanged and have continued to live with their new apparatus. CONCLUSIONS: Acoustic signal monitoring can successfully identify HeartMate XVE device end-of-life. This new method provides a low-cost, reproducible, non-invasive technique that may be used to identify possible impending device failure.
Authors: Beren Semiz; Sinan Hersek; Maziyar Baran Pouyan; Cynthia Partida; Leticia Blazquez-Arroyo; Van Selby; Georg Wieselthaler; James M Rehg; Liviu Klein; Omer T Inan Journal: IEEE J Biomed Health Inform Date: 2020-01-13 Impact factor: 5.772