Skip to main content
EJNMMI Reports

Table 2 Contingency analysis of ANN and two observers

From: Ability of artificial intelligence to diagnose coronary artery stenosis using hybrid images of coronary computed tomography angiography and myocardial perfusion SPECT

 

True findings

Infarction

Infarction + Ischemia

Ischemia

None

Total

A. Observer A

  Infarction

10

4

1

1

16

  Infarction + Ischemia

2

4

6

0

12

  Ischemia

0

1

6

4

11

  None

3

2

2

13

20

  Total

15

11

15

18

59

B. Observer B

  Infarction

9

3

2

3

17

  Infarction + Ischemia

5

6

7

5

23

  Ischemia

1

1

5

3

10

  None

0

1

1

7

9

  Total

15

11

15

18

59

C. ANN

  Infarction

10

0

1

2

13

  Infarction + Ischemia

3

9

5

0

17

  Ischemia

0

0

7

1

8

  None

2

2

2

15

21

  Total

15

11

15

18

59

D. Statistical significance

 

Likelihood ratio

p value

Pearson statistics

p value

 

  ANN

65.6

< 0.0001

68.1

< 0.0001

 

  Observer A

44.8

< 0.0001

41.1

< 0.0001

 

  Observer B

22.1

0.0087

22.4

0.0076

 
  1. The likelihood ratios, Pearson statistics, and p values for observer A (A), observer B (B), and ANN (C)
  2. Likelihood ratios, Pearson statistics, and probability (p) estimates for observers A, B, and artificial neural network (ANN)
Back to article page