The overall goal of this procedure is to assess global left ventricular diastolic function in a patient using the parametrized diastolic filling or PDF formalism. This is accomplished by first collecting data on diastolic function from a complete echocardiographic study of heart function and transferring the data to a computer for analysis. The second step is to work with the trans mitral flow images and focus on one heartbeat.
A custom program extracts conventional diastolic function parameters for the e and a waves and pre-process the image for the next step in analysis. Next, the pre-processed trans mitral flow image is loaded into a semi-automated fitting program. A segment of the EWA is selected for fitting and a preliminary parametrized diastolic filling model is produced and overlaid on the image.
The final step is to refine the fit if necessary, and have the program generate the final parametrized diastolic filling model based parameters for the input ewa. Ultimately, the process yields parameters that quantify relaxation, stiffness, and load responsible for the patient's diastolic function. These parameters have been used to distinguish between normal and pathologic function and elucidate new physiology.
The main benefit of the PDF technique is that it marries Newton's laws of motion with the actual suction pump physiology of diastole in order to extract physiologically and clinically relevant diastolic function parameters from the waveform itself. In cardiac physiology, relaxation, and the stiffness are too well-established physiological parameters that characterize diastole. The PDF method can help answer key question in the field of cardiac physiology because it uncouple the key drivers of diastole into non-invasively measurable parameters.
The PDF method is a non-invasive technique that may provide insights into determinants of diastole that were previously only measurable by invasive means. The beauty of this method is that it applies to anything in DTE that moves, so it is equally valid for blood flow across the mitral valve. As for mitral annular tissue motion, The sonogram analysis will be done using custom MATLAB and lab view programs.
Use the custom MATLAB utility to convert the DICOM formatted images to bitmap files. Next, click on the find folder with images button to load the bitmap files into the custom MATLAB program for measuring trans mitral flow parameters. Click on show next image to review the images and then select an image with a clear trans mitral flow contour and a complete cardiac cycle.
For analysis, the loaded image has time along the horizontal axis and velocity along the vertical axis. Determine the time sampling rate by clicking on the TSR button. Use the hash marks on the image to identify two time points on the time axis that are one second apart.
Then use the crosshairs to click on the lower left of the first hash mark. To zoom in on the image on the zoomed in image, click on the left edge of the desired hash mark, and the image will zoom out. Repeat the same procedure for the second hash mark.
That is exactly one second after the first hash mark. To find the velocity sampling rate first, set the unit of velocity to be marked in the box next to the VSR button. The default is one meter per second or 100 centimeters per second.
Click on the VSR button and use the crosshairs to click on the lower left of the zero velocity point to locate V equals zero on the zoomed in image. Note where the velocity scale is displayed on the image. Click on the lowest edge of the zero velocity point and the image zooms out.
Now click to the lower left of the velocity point at the chosen value, usually 100 centimeters per second, and mark the lowest edge of the second point. To complete velocity calibration in pixels, click on the mark e and a wave button. First, you define the start and end of one cardiac cycle.
Do this by marking adjacent to our peaks of the ECG following the same procedure used for marking time and velocity points. The beat duration will be calculated based on the number of pixels and the time sampling rate. After calculating the heart rate, use physiological markers to define the start of the isovolumic relaxation interval.
Next, obtain a zoomed in view of the waves of interest by clicking above the peak of the highest wave. When done, use the crosshairs to select the doppler EWA peak point. This anchors the crosshairs.
Next, move the crosshairs to the start of the ewa. Notice that the crosshairs define a line with one end anchored at the EWA peak. Before clicking, make sure the anchored line starts at the onset of the ewa.
Now mark the end of the ewa. Move the anchored line and click at the end of the ewa. These visually determined points allow calculation of the EWA acceleration and deceleration times continue to mark the wave in a similar fashion.
Once this is done, the software will automatically generate a file with measured conventional echo parameters and cropped images containing only the selected e and a waves perform the parametrized diastolic filling analysis of the ewa. In the custom lab view program, analogous steps can be followed for the A wave. Identify and load the image file created during the conventional analysis.
The lab view program automatically selects the pixels to be fit and shows them in blue, green, and red to select pixels that better represent the wave contours. Make sure change MVE is selected under show fit. Then use the maximum velocity envelope threshold level sliders to the left of the image.
Increasing the threshold value selects pixels with higher brightness. After setting the threshold, move to the EWA start slider. To begin selecting the time ranges, position the slider so that it excludes the noise at the EWA start and the associated vertical line that intersects the up slope at about half or above of the peak velocity.
Now go to the EWA end slider and position the associated vertical line about halfway up the deceleration slope so that the usual noise close to the baseline is excluded. Note, the pixels in the selected range are shown in green. When this is done, proceed by clicking fit EWA to start the PDF fitting to see the fits toggle the show fit switch to show fit.
Noisy data may make the fit very sensitive to the chosen maximum velocity threshold. If the fit is a poor approximation to the ewa, adjust it by altering the EWA time and maximum velocity envelope threshold sliders. It is also possible to modify the EWA parameters directly.
The fit parameters are displayed on screen and values can be entered in their text boxes. They can also be changed using the up and down arrows. As the parameters are modified, the green curve representing the fit will change once a satisfactory fit is obtained.
Click on update to get the updated value of the mean square error. Once the EWA has been optimized, click done to generate plots and data files for later.Analysis. Here is a normal or pseudo normal EWA pattern, which are indistinguishable using conventional indexes.
Conventional EWA derived diastolic function parameters are shown beneath the image. The PDF parameter values are also shown. In addition, the PDF model predicted fit is overlaid in green.
It demonstrates excellent fit. This is a delayed relaxation pattern with its PDF model fit. Note that compared to normal, this EWA has lower amplitude and deceleration time.
The PDF model parameter C reflecting chamber visco elasticity or relaxation is higher than for the normal pattern. The PDF model predicted fit for this constrictive restrictive pattern is in close agreement with the clinical EWA contour. The tall narrow EWA constrictive restrictive pattern is generated by chambers with increased stiffness and decreased ejection fraction.
In this example, the PDF stiffness parameter K is higher than both the normal and delayed relaxation pattern. Once mastered, this technique can be done in one to two minute for each EWA selected for analysis if it is performed properly. While performing this procedure, it's important to remember to select relatively noise free ewas whose mitral velocity envelope follows the control of trans mitral flow.
Following this procedure, other methods like fitting tissue motion can be performed in order to answer additional questions regarding longitudinal chamber stiffness and relaxation. After its conceptual derivation and in vivo validation, these techniques have paved the way for researchers in cardiovascular physiology to quantify the transverse versus longitudinal impedance. In diastole extract a load independent index of diastolic function and fractionate the EWA deceleration time into stiffness and relaxation components.
In recent work, these techniques have been used to explore the relationship between diastolic function and vortex flow generation in healthy and disease states. After having read the article and after watching this video, you should have a good conceptual understanding of how to assess global left ventricular diastolic function via the PDF formalism.
