Non-invasive myocardial work in aortic stenosis: validation and improvement in left ventricular pressure estimation

Figure 1. Normalized pressure reference curve. (A) Invasive left ventricular pressure traces from patients with aortic stenosis. Valvular events: circle, mitral valve closure; square, aortic valve opening; diamond, aortic valve closure; triangle, mitral valve opening. (B) Traces from A are normalized (light grey curves) and averaged (blue curve) with the corresponding valvular events.

Figure 2. Reference curves and estimation models. (A) The original reference curve used previously in patients without aortic stenoisis shown together with the new aortic stenosis–specific reference curve, where phase intervals were set to the average for the patients. Notice that the pressure at AVO relative to peak pressure is substantially lower in patients with aortic stenosis. Furthermore, the reference curves differ in length/duration, reflecting increased ejection times in aortic stenosis patients. (B) Estimated LVP (all three models) compared with invasive LVP (representative patient). Estimations differ mainly around the time of AVO. Early systolic pressure overestimation with the original reference curve is corrected when AVO is adjusted vertically to diastolic pressure, illustrated in Panel 2. Arrows mark adjustment to AVO. AVC, aortic valve closure; AVO, aortic valve opening; MVC, mitral valve closure; MVO, mitral valve opening; LVP, left ventricular pressure.

Figure 3. Schematic illustration of the estimated pressure–strain loop area. (A) Pressure–strain loop ‘closed’ in diastole with a straight (blue) line from MVO to MVC. (B) Pressure–strain loop when assuming zero pressure during filling shown with vertical blue lines from MVO and MVC down to the x-axis. The blue-coloured area illustrates the small relative difference in loop area if filling pressure, in a thought scenario, increases from 10 to 20 mmHg. MVC, mitral valve closure; MVO, mitral valve opening.

Figure 4. Comparison and accuracy of LVP curve estimation. (A and B) Compare measured (continuous line) and estimated LVP (dashed line) traces and point-by-point correlation from a representative patient illustrating the difference in LVP estimations with different reference curves. (C) The agreement between estimated and measured LVP (for all the patients) with the mean difference shown as a thick bright red line. Note the improvement of early systolic pressure estimation following AVO pressure adjustment (1C vs. 2C). AVO, aortic valve opening; LVP, left ventricular pressure.

Figure 5. Correlation and agreement of MWI based on invasive and estimated LVP curves. (A and B) Correlation and agreement for MWI as represented by pressure–strain loop areas calculated with measured and estimated LVP. (C) Pressure–strain loops from a selected patient, calculated with estimated (dashed line) and measured LVP (continuous line). AVO, aortic valve opening; LVP, left ventricular pressure; STE, speckle-tracking echocardiography.

Figure 6. Difference in MWI. (A) Agreement between estimated and measured MWI for all three models as a function of time. (B) The cumulative difference in MWI estimations as a function of time. The mean value is depicted with a red line. AVC, aortic valve closure; AVO, aortic valve opening; MVC, mitral valve closure; MVO, mitral valve opening; MWI, myocardial work index.

Figure 7. MWI analysis with and without AVO pressure-adjusted reference curve. Regional analysis of MWI in two patients with severe aortic stenosis and an ischaemic region (upper panel) and a dyssynchronous region (lower panel). (A) Estimated LVP curves and recorded strain trace. (B) Strain–pressure loops with direction of systolic motion marked with arrows. (C) MWI as a function of time. Estimations with (blue-dashed line) and without (red-dashed line) AVO pressure adjustment. Regional longitudinal strain (green line). AVO, aortic valve opening; MWI, myocardial work index; LVP, left ventricular pressure.