Deep learning assisted measurement of echocardiographic left heart parameters: improvement in interobserver variability and workflow efficiency.

Machine learning techniques designed to recognize views and perform measurements are increasingly used to address the need for automation of the interpretation of echocardiographic images. The current study was designed to determine whether a recently developed and validated deep learning (DL) algorithm for automated measurements of echocardiographic parameters of left heart chamber size and function can improve the reproducibility and shorten the analysis time, compared to the conventional methodology. The DL algorithm trained to identify standard views and provide automated measurements of 20 standard parameters, was applied to images obtained in 12 randomly selected echocardiographic studies. The resultant measurements were reviewed and revised as necessary by 10 independent expert readers. The same readers also performed conventional manual measurements, which were averaged and used as the reference standard for the DL-assisted approach with and without the manual revisions. Inter-reader variability was quantified using coefficients of variation, which together with analysis times, were compared between the conventional reads and the DL-assisted approach. The fully automated DL measurements showed good agreement with the reference technique: Bland-Altman biases 0-14% of the measured values. Manual revisions resulted in only minor improvement in accuracy: biases 0-11%. This DL-assisted approach resulted in a 43% decrease in analysis time and less inter-reader variability than the conventional methodology: 2-3 times smaller coefficients of variation. In conclusion, DL-assisted approach to analysis of echocardiographic images can provide accurate left heart measurements with the added benefits of improved reproducibility and time savings, compared to conventional methodology.

Normal Values of Cardiac Output and Stroke Volume According to Measurement Technique, Age, Sex, and Ethnicity: Results of the World Alliance of Societies of Echocardiography Study.

BACKGROUND: Assessment of cardiac output (CO) and stroke volume (SV) is essential to understand cardiac function and hemodynamics. These parameters can be examined using three echocardiographic techniques (pulsed-wave Doppler, two-dimensional [2D], and three-dimensional [3D]). Whether these methods can be used interchangeably is unclear. The influence of age, sex, and ethnicity on CO and SV has also not been examined in depth. In this report from the World Alliance of Societies of Echocardiography Normal Values Study, the authors compare CO and SV in healthy adults according to age, sex, ethnicity, and measurement techniques. METHODS: A total of 1,450 adult subjects (53% men) free of heart, lung, and kidney disease were prospectively enrolled in 15 countries, with even distributions among age groups and sex. Subjects were divided into three age groups (young, 18-40 years; middle aged, 41-65 years; and old, >65 years) and three main racial groups (whites, blacks, and Asians). CO and SV were indexed (cardiac index [CI] and SV index [SVI], respectively) to body surface area and height and measured using three echocardiographic methods: Doppler, 2D, and 3D. Images were analyzed at two core laboratories (one each for 2D and 3D). RESULTS: CI and SVI were significantly lower by 2D compared with both Doppler and 3D methods in both sexes. SVI was significantly lower in women than men by all three methods, while CI differed only by 2D. SVI decreased with aging by all three techniques, whereas CI declined only with 2D and 3D. CO and SV were smallest in Asians and largest in whites, and the differences persisted after normalization for body surface area. CONCLUSIONS: The present results provide normal reference values for CO and SV, which differ by age, sex, and race. Furthermore, CI and SVI measurements by the different echocardiographic techniques are not interchangeable. All these factors need to be taken into account when evaluating cardiac function and hemodynamics in individual patients.