7.2 Strain Acquisition, Tracking Quality, and Reproducibility
Key Takeaways
- GLS measures longitudinal myocardial deformation and conventionally retains a negative sign for systolic shortening.
- Nonforeshortened A4C, A2C, and apical long-axis loops, complete myocardium, useful frame rate, and clean ECG are prerequisites for reliable tracking.
- Every segment must be visually checked through the cardiac cycle because smoothing, curves, and polar maps can conceal tracking of cavity, pericardium, or artifact.
- Serial comparison is strongest with matched physiology and acquisition, the same vendor and software release, saved raw loops, and documented exclusions and limitations.
Strain begins with deformation and convention
Strain describes change in length relative to the original length. Longitudinal systolic shortening is conventionally negative, so a segment that shortens from its end-diastolic length has negative longitudinal strain. Global longitudinal strain, or GLS, summarizes longitudinal deformation across the analyzed LV segments; it is not ejection fraction expressed with another unit. Retain the negative sign in display and reporting unless the laboratory explicitly reports the absolute magnitude and labels it as global longitudinal shortening. A value that is less negative represents less shortening, but interpretation requires loading, rhythm, image quality, software, and clinical context.
Speckle-tracking echocardiography follows acoustic patterns frame to frame. It is angle independent in principle but not image independent: dropout, reverberation, lateral resolution, out-of-plane motion, and foreshortening can make software follow the wrong feature. Current ASE guidance favors two-dimensional speckle tracking for comprehensive clinical LV GLS. Tissue-Doppler-derived strain remains more angle dependent and is not interchangeable with speckle-tracking GLS.
Acquire comparable apical loops
Acquire nonforeshortened apical four-chamber, two-chamber, and long-axis views that include the entire LV myocardium. The true apex should be beyond the cavity's long axis, not a rounded cap created by an overly high window. Keep depth just sufficient for the LV, narrow the sector without cutting off myocardium, center the ventricle, optimize focus and frequency, and balance gain so endocardium is visible without filling the wall with noise. Record a clean ECG because beat selection and valve-event timing depend on it.
Use the laboratory and vendor protocol for frame rate. A larger sector and greater depth reduce frame rate; an extremely high rate obtained by sacrificing spatial information may also impair tracking. Save native tissue loops before contrast when the protocol requires it, because contrast interfaces can alter tracking. Record heart rate, rhythm, blood pressure, and relevant loading conditions. For serial studies, reproduce patient position, view geometry, acquisition settings, beat selection, and physiologic conditions as closely as practical.
| Acquisition problem | What it does | Best technical response |
|---|---|---|
| Foreshortened apex | Changes LV length and segment deformation | Reposition to the true apical window and reacquire all required views |
| Endocardial dropout | Invites tracking into cavity, pericardium, or artifact | Adjust window, frequency, gain, and focus; use another interspace if needed |
| Low frame rate | Misses rapid motion and reduces temporal tracking | Reduce depth or sector width while retaining the entire myocardium |
| Very narrow sector that clips wall | Excludes required segments despite a higher rate | Widen enough to include the complete myocardial ROI |
| Ectopy or variable cycle length | Produces beat-dependent strain and timing | Select protocol-defined representative beats and document the rhythm limitation |
In atrial fibrillation, a single attractive loop may not be representative. Use the laboratory's validated beat-selection or averaging method, seek comparable preceding and pre-preceding intervals when applicable, and document variability. Do not substitute a postectopic beat merely because its border is clearer; altered filling and contractility change deformation.
Verify tracking rather than trusting the bull's-eye
Select end-diastole and end-systole consistently. Aortic-valve closure can be derived from the apical long-axis view, Doppler, or a validated automatic method, but the same method should support comparison. Trace the endocardial contour at the software-defined reference frame, then adjust the region of interest to contain the full myocardial wall while excluding blood pool, pericardium, papillary muscles, and adjacent structures. Current ASE guidance generally reports midwall strain.
Play the tracking overlay through the entire cardiac cycle and compare it directly with myocardial motion in every segment. A smooth curve or complete polar map does not prove validity; software smoothing can hide tracking failure. If a basal segment follows pericardium, the contour jumps into the cavity, or the apex slips out of plane, correct the contour and region. Reacquire when image information is insufficient. Reject an uncorrectable segment according to the validated software and laboratory rule rather than accepting it to make the map complete.
GLS is a global measure, while regional segment values and patterns are less reproducible and more vulnerable to local artifact. Review the curves, timing, tracking overlay, and source loops before accepting the number. Dyskinetic lengthening may generate positive systolic strain; do not remove the sign merely to make all segments look alike. The sonographer documents the acquisition and analysis quality, while the interpreting physician integrates the result with function, loading, and disease.
Design serial studies for reproducibility
Use the same vendor and, ideally, the same software release for serial measurements. Vendors have improved agreement for global values, but algorithms, border conventions, and regional results can still differ. An apparent change after an upgrade, different platform, major blood-pressure change, or different rhythm is not automatically biologic. Save raw loops, record the platform and analysis version when available, and state excluded segments or limited views.
For surveillance, first compare technical comparability: views, frame rate, rhythm, loading, timing convention, ROI, tracking acceptance, and software. Only then compare GLS. If a patient's blood pressure is markedly higher and two segments track poorly, reacquisition and transparent limitation are more defensible than reporting a precise delta. Reproducibility is produced by a controlled workflow, not by copying the prior value or adjusting contours until the new result matches it.
Watch the myocardium, not just the curves
Review the tracking overlay through the full cardiac cycle in every view. A smooth curve and colorful polar map can still be generated when the contour follows cavity, pericardium, or out-of-plane artifact.
During GLS analysis, the basal lateral region of interest follows the pericardium for part of systole even though the software displays a smooth curve. What is the best action?
Which three practices improve comparability in a serial LV GLS study? Select three correct responses.
Select all that apply