A graphical analysis of aspects contributing to the spreading of measurements of left ventricular function

Figures

• 

  Fig. 1. a–e Graphical analysis of measurement deviation on a 4-chamber view. All cardiologists were asked to draw contours for 15 four chamber view on an end-diastolic and end-systolic image. Then, all individual contours were combined on one echocardiogram. Also, the reference contour of the group was calculated. The images display the following: a 4-chamber view individual end-systolic contours (pink are the individual contours, blue is the reference contour), b 4-chamber view with depiction of distance between individual contours and reference contour per segment regarding end-systolic contours, c positioning of the posterior mitral valve ring, d positioning of the apex, e positioning of the anterior mitral valve ring

• 

  Fig. 2. Depiction of the median distance between reference and individual contour per segment. For left ventricular segments, the distance [mm] (median ± IQR) between individual contour and reference contour was depicted. This was done for 2-chamber view (upper panel) and 4-chamber view (lower panel), for end-diastole and end-systole

• 

  Fig. 3. a Graphical depiction of the spreading of distinct anatomical landmarks. This figure displays the individual location of the apical and mitral valve contour (both, end-diastolic and end-systolic and 4-chamber view (left) and 2-chamber view (right)). For illustration, the reference position of all measurements is defined and the distance [mm] of the individual measurements calculated from that point (x-axis: horizontal image plane, y-axis: vertical image plane). The inner ring covers 1cm2, the following 4cm2 and the outer ring 9cm2. Image quality is also displayed to illustrate the influence on landmark position: green = excellent, yellow = good, orange = fair. b Violin plot of the spreading of distinct anatomical landmarks. This figure shows violin plots (depiction the distribution of measurements and markers for the median/ interquartile range) for the distance [cm] between reference position and individual location of the mitral valve ring and apical contour (both, end-systolic/ left and end-diastolic/ right)

• 

  Fig. 4. Depiction of the left ventricular ejection fraction based on Simpson measurement. The left ventricular ejection fraction measured using Simpson’s method on echocardiogram was displayed (box plots indicating: median with 25th/75th percentiles). This was done per a echocardiogram (1–15) and b per cardiologist (1–42)

• 

  Fig. 5. a Depiction of left ventricular end-diastolic and end-systolic by different cardiologists. This figure displays the difference between left ventricular volumes measured by individual cardiologists and the median of the group. The x-axis depicts the individual cardiologist, the y-axis the percentage of underestimation (negative deflection) or overestimation (positive deflection). Light red/ light grey shows the difference in end-diastolic volume (combining 2-chamber view and 4-chamber view), dark red/ dark grey depicts end-systolic volume (again, combining 2-chamber view and 4-chamber view). b Depiction of the left ventricular ejection fraction depending on tracing behavior. The relation between left ventricular (LV) ejection fraction and the location of the contours compared to the reference contour of the group is displayed. All individual cardiologists were allocated to three different groups based on the localization of their drawn contours: (1) small contours of the LV compared to the reference contour of the group (n = 13), (2) contours in line with the reference contour of the group (n = 21) or (3) contours larger than the reference contour of the group (n = 8). Left Panel:…

• 

  Fig. 6. Depiction of all apical contours of the left ventricle in a three-dimensional model. This figure displays the localization of the apex based on individual positioning in three dimensions. More specifically, the apical localizations in the 2-chamber view and 4-chamber view were registered to determine a three-dimensional position. The length of the different left ventricle were standardized and superimposed on a heart model. Then, the apical position was noted (as yellow dots) on that cast