Skip to main content

Table 1 Subjective image quality analysis

From: Deep learning image reconstruction algorithm for carotid dual-energy computed tomography angiography: evaluation of image quality and diagnostic performance

 

ASIR

DLIR-L

DLIR-M

DLIR

p value

κ values

Image noise

Reader 1

3.4 ± 0.5 (3–4)

3.3 ± 0.6 (2–4)

3.7 ± 0.4* (3–4)

4.5 ± 0.5* (4–5)

0.000

0.53

Reader 2

3.5 ± 0.5 (3–4)

3.5 ± 0.6 (2–4)

3.7 ± 0.4* (3–4)

4.4 ± 0.5* (4–5)

0.000

 

Image texture

Reader 1

3.5 ± 0.6 (2–4)

3.8 ± 0.5* (3–5)

3.8 ± 0.4* (3–5)

4.4 ± 0.6* (3–5)

0.000

0.57

Reader 2

3.5 ± 0.6 (2–4)

3.7 ± 0.4* (3–4)

3.8 ± 0.4* (3–4)

4.6 ± 0.6* (3–5)

0.000

 

Overall image quality

Reader 1

3.4 ± 0.4 (3–4)

3.6 ± 0.4 (3–5)

3.8 ± 0.3* (3–5)

4.4 ± 0.3* (3–5)

0.000

0.44

Reader 2

3.5 ± 0.3 (3–4)

3.6 ± 0.3 (3–4)

3.8 ± 0.3* (3–4)

4.4 ± 0.3* (4–5)

0.000

 
  1. *Mean statistical higher than ASIR-V
  2. ASIR, adaptive statistical iterative reconstruction-Veo; DLIR, deep learning image reconstruction (low strength, DLIR-L; medium strength, DLIR-M; high strength, DLIR-H)