Usefulness of automatisation in measurement of left ventricular global longitudinal strain for beginners in echocardiography

Background/Introduction:
Global longitudinal strain (GLS) offers valuable insight into left ventricular (LV) function beyond ejection fraction in echocardiography (Echo) but is underutilised due to expertise and time constraints.

Purpose:
We aimed to assess time and agreement in GLS measurement by a beginner cardiologist compared to an expert using semi-automated (Semi-Auto) and conventional (Manual) methods in speckle tracking analysis.

Methods:
The Semi-Auto method requires only manual selection of apical chamber views (CV) for analysis and performs automatic classification, endocardial border tracing, and GLS calculation of apical CV with minimal involvement for tracking adjustments. In contrast, the Manual method necessitates full user involvement, relying solely on the operator's expertise in all steps. User involvement time included CV selection, setting endocardial border points (systole and diastole), and tracking adjustments. Measurements were stored separately for final analysis against those of the expert cardiologist. The beginner's training involved 10 exams using Semi-Auto and Manual methods. Then, 90 random Echo exams were analysed sequentially by the beginner first using Semi-Auto, followed by Manual after a three-week interval to the same sequence of exams to eliminate bias.

Results:
As shown in Figure 1 (top row), the beginner's involvement time for Semi-Auto remained low and stable, while Manual showed an inverse sigmoid curve converging at the 50th analysis. Higher correlation (r = 0.85) was achieved with Semi-Auto GLS between beginner and expert than with Manual (r = 0.74) (Figure 1, middle row). Concordantly, GLS measurement bias was more pronounced with the Manual method between beginner and expert for the entire bout than with the Semi-Auto method (Figure 1, bottom row).

Conclusion:
The findings highlight the potential of automated GLS analysis to bridge the expertise gap and streamline cardiac diagnostics. Integrating automated solutions into routine clinical practice could enhance efficiency and accuracy, particularly for beginners or infrequent users. This advancement paves the way for more standardised and reliable GLS measurements.