Table 2 Comparison of the group mean ± standard deviation of morphological metrics between measurements performed using manual landmarks on CT images and measurements performed on MRI using our proposed methods.

From: Deep learning algorithms enable MRI-based scapular morphology analysis with values comparable to CT-based assessments

 

CT

MRI

p-value

Diff (abs)

ICC

Mean

SD

Mean

SD

mean

SD

 

95% CI

Interpretation

Glenoid radius [mm]

12.6

1.2

12.0

±1.2

0.03

0.8

±0.5

0.8

(0.40–0.95)

Good (95%CI poor-excellent)

Glenoid height [mm]

36.4

3.7

36.1

±4.3

0.85

1.7

±0.8

0.9

(0.62–0.97)

Good (95%CI moderate-excellent)

Glenoid version [°]

0.0

5.11

-1.5

±4.2

0.38

2.6

±2.5

0.7

(0.23–0.93)

Moderate (95%CI poor-excellent)

Critical shoulder angle [°]

33.3

4.26

32.0

±4.0

0.04

1.8

±1.1

0.9

(0.69–0.98)

Excellent (95%CI moderate-excellent)

Glenoid inclination [°]

-6.4

5.69

-5.2

±5.2

0.16

1.6

±1.8

0.9

(0.74–0.98)

Excellent (95%CI moderate-excellent)

  1. The differences of the values form CT and MRI are analysed using the p-value of the wilcoxon test between the two measurements as well as mean and standard deviation of the absolute differences for each parameter. The inter class correlation (ICC) between the measurements from CT and MRI are presented. Intraclass correlation (ICC), confident interval (CI).
Back to article page