Author + information
- Kegan Moneghetti,
- Juyong Kim,
- Genevieve Giraldeau,
- Yukari Kobayashi,
- Kalyani Boralkar,
- Ingela Schnittger,
- Euan Ashley,
- Matthew Wheeler and
- Francois Haddad
Background: Advances in echocardiographic acquisition and analysis has led to an array of reported variables beyond time honored left ventricular ejection fraction (LVEF). Selecting a combination of variables which provide the greatest contribution to risk prediction remains a challenge.
Methods: Two hundred and one patients with dilated cardiomyopathy (mean LVEF 33±12%) and comprehensive echocardiography metrics were selected and followed for a median of 4.5 years for death, heart transplantation and left ventricular device insertion. Correlation mapping was used to discover associations within the set of echocardiographic variables. R2 values were used as distance between variables and iterated for 400 times until the system was in dynamic equilibrium to generate a correlation map (Figure 1).
Results: Of the metrics of LV function, LVEF and LV global longitudinal strain had the strongest correlation (r = −0.89, p<0.01) and together appeared strongly correlated to LV strain rate and LV end systolic diameter. Age was associated with E'. Right atrial volume index (RAVI) appeared the least associated variable to the cluster of LV function metrics. When RAVI was placed in a proportional hazard model with LVEF it provided incremental prognostic benefit when compared to a combination of variables clustered with LVEF (x2 = 43 vs 36, p <0.01).
Conclusions: Correlation mapping appears to provide a framework for selection of the optimal echocardiographic variables to enter into risk prediction models.
Poster Hall, Hall C
Saturday, March 18, 2017, 9:45 a.m.-10:30 a.m.
Session Title: Innovative Use of Echocardiography
Abstract Category: 28. Non Invasive Imaging: Echo
Presentation Number: 1197-216
- 2017 American College of Cardiology Foundation