• H Shih, Z Hillel, C Declerck, C Anagnostopoulos, M Kuroda, and D Thys.
• St. Luke's-Roosevelt Hospital Center, Columbia University College of Physicians and Surgeons, New York, NY 10025, USA.
• J Clin Monit. 1997 May 1; 13 (3): 157-70.

AbstractWe describe a computer algorithm that allows continuous, real-time evaluation of ventricular elastance (Ees), arterial elastance (Ea), and their coupling ratio in a clinical setting. In the conventional pressure-volume analysis of left ventricular (LV) contractility, invasive methods of volume determination and a significant, rapid preload reduction are required to generate Ees. With the help of automated border detection by transesophageal echocardiography, and a technique of estimating peak LV isovolumic pressure, Ea and Ees were determined from a single cardiac beat without the need for preload reduction. A comparison of results obtained by a conventional approach and the new algorithm technique, showed good correlation for Ea (r = 0.86, p < 0.001) and Ees (r = 0.74, p = 0.001). Bias analysis showed a bias (d) of 1.47 mmHg/cm2 for Ea with a standard deviation (SD) of 7.03 mmHg/cm2, and upper (d+2SD) and lower(d-2SD) limits of agreement of 15.24 mmHg/cm2 and -12.31 mmHg/cm2, respectively. Bias analysis showed a bias of -1.42 mmHg/cm2 for Ees with a SD of 4.88 mmHg/cm2, and limits of agreement of 8.15 mmHg/cm2 and -10.98 mmHg/cm2. The algorithm's stability to artifacts was also analyzed by comparing magnitudes of residuals of Ea and Ees from source signals with and without noise. With Ea differing by an average of 1.036 mmHg/cm2 and Ees differing by an average of 0.836 mmHg/cm2, the algorithm was found to be stable to artifacts in the source signals.

12 keywords...

hide…