- Poster presentation
- Open Access
Principle component analysis of myocardial strain to quantify left ventricular dyssynchrony
© Ramachandran et al; licensee BioMed Central Ltd. 2013
- Published: 30 January 2013
- Cardiac Resynchronization Therapy
- Mechanical Dyssynchrony
- Myocardial Strain
- Left Ventricular Dyssynchrony
- Left Ventricular Segment
Cardiac resynchronization therapy (CRT) is effective for selected heart failure (HF) patients, but is associated with a 30-40% nonresponse rate. Identification of CRT responders may be improved with myocardial strain imaging. The circumferential uniformity ratio estimate (CURE )1 measures mechanical dyssynchrony by Fourier series fitting of myocardial strains over space, but requires user interaction to define a range of cardiac phases over which CURE is calculated (time dependence). We hypothesize that principal component analysis (PCA) can quantify dyssynchrony in myocardial strain in a data-driven, time-independent manner that does not require any subjective user assessments of strain data.
Dyssynchronous HF was induced in canines (N=5) with tachycardia pacing and left bundle branch ablation (LBBB-HF), while synchronous HF with narrow QRS (NQRS-HF) was induced in canines with tachycardia pacing only (N=5). Four normal canines were also studied.
Spiral cine DENSE MRI was performed on a 1.5 T scanner (Avanto, Siemens) in all 14 canines. 2D myocardial motion was tracked in a mid-ventricular short-axis plane from DENSE images acquired using the following parameters2: interleaves=6, TR=17 ms, TE=1.9 ms, slice thickness=8 mm, excitation flip angle=15°, in-plane resolution=2.8 x2.8 mm and displacement-encoding frequency= 0.1 cycles/mm.
DENSE images were analyzed3 to calculate left ventricular (LV) circumferential strain (Ecc), and PCA was applied to the Ecc-time curves for a 24-segment LV model. Specifically, the LV Ecc curves were decomposed spatially into principal component basis vectors. The PCA-based metric for measuring LV dyssynchrony, termed First Order Regional Conformity Estimate (FORCE), was calculated as |sum(PCL1)|/sum(|PCL1|) where PCL1 represents the loadings of the first principal component basis vector. Both FORCE and CURE range from 0 (dyssynchrony) to 1 (perfect synchrony). The Kruskal-Wallis one-way ANOVA test was used to compare FORCE and CURE among the three groups in pairwise fashion.
PCA, using the promising new FORCE parameter, effectively and automatically identifies mechanical dyssynchrony in HF in a data-driven and completely time-independent fashion. Further clinical evaluation of FORCE for prediction of CRT response is warranted.
This work was funded by the American Heart Association 12GRNT12050301 and NIH Heart, Lung and Blood Institute (NHLBI) T32HL007284.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.