Skip to main content

Table 3 Inter-observer variability in three observers with different level of expertise

From: Cardiovascular magnetic resonance myocardial feature tracking using a non-rigid, elastic image registration algorithm: assessment of variability in a real-life clinical setting

 

Inter-observer

#

Expert versus Skilled Observer

Expert versus Beginner

 

Bias

LOA

CV

CR

ICC (95% CI)

Bias

LOA

CV

CR

ICC (95% CI)

Healthy Volunteer

Global

ErrSAX

120

0.01

[−3.3;3.4]

4.3

±4.8

0.97 (0.97–0.98)

−0.61*

[−6.8;5.6]

8.3

±8.8

0.99 (0.98–0.99)

EccSAX

120

0.46*

[−1.5;2.5]

5.4

±2.9

0.98 (0.98–0.98)

−0.34*

[−2.9;2.3]

6.9

±3.7

0.98 (0.97–0.99)

EllLAX

20

0.50*

[−0.7;1.7]£

4.0

±1.7

0.85 (0.46–0.95)

0.70*

[−2.4;3.8]£

5.3

±2.2

0.74 (0.45–0.89)

Segm.

ErrSAX

660

0.01

[−8.8;8.8]£

16.4

±12.4

0.96 (0.96–0.96)

−0.43

[−11.8;10.9]£

22.9

±16.0

0.95 (0.94–0.95)

EccSAX

660

0.55*

[−5.6;6.7]£

17.1

±8.7

0.95 (0.94–0.96)

−0.22

[−10.2;9.8]£

22.7

±14.2

0.92 (0.91–0.93)

EllLAX

140

0.60*

[−5.3;6.5]£

18.7

±8.3

0.90 (0.86–0.93)

0.86*

[−8.4;10.1]£

19.6

±13.1

0.80 (0.76–0.90)

Normal Group

Global

ErrSAX

87

0.57*

[−2.7;3.8]

5.7

±4.6

0.97 (0.97–0.97)

0.19 *

[−3.4;3.8]

7.3

±5.1

0.97 (0.96–0.97)

EccSAX

87

0.56*

[−1.3;2.4]

5.2

±2.6

0.99 (0.98–0.99)

−0.16

[−1.8;1.4]

5.0

±2.2

0.99 (0.98–0.99)

EllLAX

20

−0.03

[−1.5;1.5]

4.4

±2.2

0.89 (0.73–0.95)

0.29*

[−1.4;2.0]

5.2

±2.4

0.91 (0.87–0.96)

Segm.

ErrSAX

470

0.57*

[−8.8;9.9]§£

21.3

±13.2

0.92 (0.89–0.95)

0.14

[−10.6;10.9]§£

21.8

±15.2

0.92 (0.91–0.94)

EccSAX

470

0.55*

[−4.4;5.5]

15.0

±7.0

0.96 (0.96–0.97)

−0.18

[−5.5;5.1]£

17.3

±7.5

0.94 (0.93–0.95)

EllLAX

140

−0.17

[−4.5;4.2]

19.4

±6.2

0.89 (0.85–0.92)

0.18

[−5.2;5.5]§

19.7

±7.5

0.90 (0.86–0.92)

Patient Group

Global

ErrSAX

86

0.26

[−3.2;3.7]

8.6

±4.9

0.97 (0.97–0.99)

0.09

[−2.8;3.0]

6.7

±4.1

0.99 (0.98–0.99)

EccSAX

86

0.51*

[−1.0;2.0]

6.0

±2.2

0.99 (0.96–0.99)

0.13

[−1.5;1.7]

6.2

±2.2

0.98 (0.97–0.99)

EllLAX

18

0.71*

[−1.2;2.7]

7.9

±3.7

0.97 (0.85–0.99)

−0.15

[−2.5;2.2]

7.9

±3.3

0.97 (0.88–0.99)

Segm.

ErrSAX

470

0.26

[−8.2;8.7]§£

25.2

±11.9

0.93 (0.90–0.95)

0.13

[−8.6;8.9]§£

23.9

±12.4

0.92 (0.88-0.95)

EccSAX

470

0.63*

[−4.5;5.8]£

19.4

±7.3

0.95 (0.94–0.96)

0.14

[−4.9;5.2]£

21.0

±7.2

0.95 (0.93-0.96)

EllLAX

126

0.65*

[−5.7;7.0]

24.8

±9.0

0.92 (0.88–0.94)

0.44*

[−6.1;7.0]§

25.7

±9.3

0.90 (0.88–0.94)

  1. # - Sample Size, LOA Limits of agreement; CV Coefficient of variation (%); CR Coefficient of repeatability; ICC Intraclass correlation coefficient; CI Coefficient Interval
  2. *p < 0.05, two tailed paired t-test against zero
  3. § p < 0.05, two tailed F-test between patient group and normal group for each strain component
  4. £ p < 0.05, two tailed F-test between the results obtained by the different observers (expert vs skilled observer, and expert versus beginner)
Back to article page

Journal of Cardiovascular Magnetic Resonance

ISSN: 1532-429X

Contact us