Table 3 (Error analysis) Present the analysis of misclassified exams.

From: Automatic diagnosis of the 12-lead ECG using a deep neural network

 

DNN

cardio.

emerg.

stud.

 

meas.

noise

unexplain.

meas.

noise

concep.

atte.

meas.

noise

concep.

atte.

meas.

noise

concep.

atte.

1dAVb

3

2

1

8

3

  

15

3

  

13

3

3

 

RBBB

3

 

1

4

 

2

 

1

 

8

 

3

 

2

 

LBBB

   

1

1

1

  

1

4

  

2

3

 

SB

4

  

4

   

4

  

1

5

 

2

1

AF

 

2

1

 

4

2

  

2

5

  

3

7

 

ST

2

 

1

2

1

 

5

1

1

1

1

1

2

1

5

  1. The errors were classified into the following categories: (i) measurements errors (meas.) were ECG interval measurements preclude the given diagnosis by its textbook definition; (ii) errors due to noise, where we believe that the analyst or the DNN failed due to a lower than usual signal quality; and (iii) other type of errors (unexplain.). Those were further divided, for the medical residents and students, into two categories: conceptual errors (concep.), where our reviewer suggested that the doctor failed to understand the definitions of each abnormality, and attention errors (atte.), where we believe the error could be avoided if the reviewer had been more careful.
Back to article page