EF accuracy ECHO v/s MRI

[aroran412]

Hello,
I am a 39 year old male, was dxed with cardiomyopathy and CHF last summer with an EF of 25% and very aggressive symptoms.  Over the course of several months, my functioning improved drastically and I have been virtually free of symptoms, suggestive of  viral myocarditis event.  In Nov I had a cardiac MRI which showed that the cardiomyopahty was under control and the MRI gave an EF of 53%.  I ushered in the new year with joy.  Have been feeling wonderful since then.  However an ECHO conducted last week shows the EF to be 30-35%.  I am totally confused and in shock.  I feel 90-95% of my former self yet the ECHO says something else.  Can anyone explain this discrepancy between the MRI and Echo.  Are these apples and oranges? the MRI folks tell me their method is more accurate while the ECHO guys say the reverse.

Any insights would be appreciated.

Thanks

[kenkeith]

To answer your questions requires some documentation which echo (2D or 3D).  Whether the procedure was manual tracing (border outlining) or automated with installed software.  Manual tracing of LV borders on 2-D is subjective, experience dependant, tedious and time consuming, and the other method that is visual is inherently subjective and convenient (Visual is probably the most used method).  Also, whether heart's inside chamber is contoured, malformed, etc
.
To begin to differentiate with an evaluation 3D echo was always superior to 2D echo regardless of sonographer experience. Three-dimensional echocardiography allows even novice sonographers to obtain diagnostic-quality data sets, which they were unable to accomplish with 2DE.

Quantification of LVV by 2-dimensional echocardiography is limited by its heterogeneous geometry, which restricts application of simple geometric models that assume symmetrical shape. Often the inside of chamber is contoured so that is another consideration.

"...Three-dimensional echocardiography provides an accurate measurement of LVV and function by the reconstruction of the true left ventricular geometry, but it remains time consuming. This study was performed to validate the 3DE-B method as a rapid method for LVV and EF calculation against the standard of MRI. Our results demonstrate an excellent correlation and close limits of agreement between LVV and EF measured by 3-dimensional echocardiography using 2 different methods (A and B) and MRI for all the subjects recruited in this study as well as for the subgroups (Table IV). Thus with the rapid 8 parallel equidistant slice method we have achieved at least the same accuracy as with the time-consuming 2.9-mm fixed slice thickness method. Clinical utility of automated assessment of left ventricular ejection fraction using artificial intelligence-assisted border detection...".

The clinical study showed superior performance of the deformable model in assessing ejection fraction when compared to MRI measures. It also showed that the three-dimensional deformable model improved EF measures, which is explained by a more accurate segmentation of small and convoluted ventricular shapes when integrating the third spatial dimension.