“Strain” in everyday language can mean “stretching” and it is used to describe “deformation” (D’Hooge 2000). The MyoStrain report will show the amount of strain (deformation) for 37 segments of the left ventricle and 11 segments of the right ventricle. The value of each segment is calculated by taking the average of “peak strain” of each of the pixels within that segment.
In a heathy heart, the wall will deform well as it goes through cardiac cycle; however, in an unhealthy heart, the heart wall might be stiffer and not move or “deform” as much. Or even worse, the muscle can be so weak that the wall bulges out instead of compressing as the heart goes into systolic phase. The strain numbers are negative because we are measuring compression of the myocardium. A lower strain number indicates more compression of the given region of the myocardium.
More information regarding the strain legend used in the MyoStrain program can be found from the following papers:
Neizel M, et al. “Strain-encoded MRI for evaluation of left ventricular function and transmurality in acute myocardial infarction.” Circ Cardiovasc Imaging. 2009;2(2):116-122 View Publication >>
Wong DT, et al. “Magnetic resonance-derived circumferential strain provides a superior and incremental assessment of improvement in contractile function in patients early after ST-segment elevation myocardial infarction.” European Radiology. 2014;24:1219-1228. View Publication >>
Oyama-Manabe N, et al. “Identification and further differentiation of subendocardial and transmural myocardial infarction by fast strain-encoded (SENC) magnetic resonance imaging at 3.0 Tesla” European Radiology. 2011;21(11):2362-2368. View Publication >>
Neizel M, et al. “Impact of Systolic and Diastolic Deformation Indexes Assessed by Strain-Encoded Imaging to Predict Persistent Severe Myocardial Dysfunction in Patients After Acute Myocardial Infarction at Follow-Up.” Journal of the American College of Cardiology. 2010;56:1056-1062. View Publication >>
Choi E-Y, et al. “Prognostic value of myocardial circumferential strain for incident heart failure and cardiovascular events in asymptomatic individuals: the Multi-Ethnic Study of Atherosclerosis.” European Heart Journal. 2013;34:2354-2361. View Publication >>
Koos R, et al. “Layer-specific strain-encoded MRI for the evaluation of left ventricular function and infarct transmurality in patients with chronic coronary artery disease.” Int J Cardiol. 2013;166:85-89. View Publication >>
The AHA models used in MyoStrain are derived from the following publication:
Cerqueira et al., “Standardize Myocardial Segmentation and Nomenclature for Tomographic Imaging of the Heart,” Circulation, 2002;105:539-542 View Publication >>
Normal Ranges of MyoStrain Measurements
The output of the SENC images post-processing is a report that shows various measurements. One set of measurements is the traditional global measurements (ejection fraction, chamber volumes and masses).
These measurements are presented with the normal ranges published by Zhan et al.¹ The other set is the strain measurements (circumferential and longitudinal) presented with the normal ranges of strains as published by Neizel et al.²
Traditional Global measurements (LVEF and indexed LVEDV, LVESV, LVSV and LV Mass):
1. Y. Zhan et al., “Derivation of consolidated normal reference values for right and left ventricular quantification by cardiac magnetic resonance using a novel meta-analytic approach,” Journal of Cardiovascular Magnetic Resonance, vol. 18, no. 1, p. O75, 2016/01/27 2016 View Publication >>
Strain (circumferential and longitudinal):
2. M Neizel et al. “Strain-encoded MRI for evaluation of left ventricular function and transmurality in acute myocardial infarction.” Circ Cardiovasc Imaging. 2009;2(2):116-122. View Publication >>
The accuracy of MyoStrain measurements are determined by the Limit of Agreement (LOA). The LOA is the range that covers the 95% of differences between the measurement of the two devices. For example, if MyoStrain measures LVEF of 67%, the LOA of (-13,+10) means that 95% for a large number LVEF values of the same subject measured using the gold standard Cine MRI will lie between 54% and 77%. The LOA depends on many factors, including images quality and inter-operator and inter-observer variabilities. This LOA was originally based on two predicate devices (Diagnosoft HARP for Strain, Diagnosoft VIRTUE for traditional measurements), however these ranges have been reduced to reflect measured accuracy in MyoStrain.
To demonstrate the accuracy of MyoStrain measurements, we calculated the correlation coefficients using Diagnosoft VIRTUE 5.51 measurements of the LV EF, End-Diastolic Volume (EDV), End-Systolic Volume (ESV), Mass, and Stroke Volume (SV). We considered the measurements accurate by requiring that the variations in global measurements due to workstation and user variability to be within the accepted cutoffs of published guidelines and clinical results. We specified that targeted correlation coefficients of the global measures generated by MyoStrain, in comparison to the gold standard cardiac MRI, must be equal or better than the following: EF: R=0.79, p<0.001, EDV: R=0.84, p<0.001, ESV: R=0.94, p<0.001, SV: R=0.31, p=0.05.
Since the LV Mass is similar to LV volumes, we decided that the R and p values follow the same criteria for EDV. Our acceptance criteria required the following bounds for the 95% range of measurements differences between MyoStrain and Diagnosoft VIRTUE:
Based on a sample size N=23 of healthy subjects and patients, MyoStrain demonstrated the following acceptable LOA:
The Bland-Altman Graphs of these calculations can be seen below:
The LOA and accuracy of Strain calculations were based on tests using a mechanical phantom with known actual strain values. Phantom analysis demonstrated that MyoStrain has the acceptable LOA of (-5,+5). Note that LOA of MyoStrain of Strain measured in vivo in humans is unknown and could be different.