Skip to main content

Table 2 ECG record description for three-class case (N, AF, and Non-AF)

From: AFibNet: an implementation of atrial fibrillation detection with convolutional neural network

Dataset

Frequency sampling (Hz)

Conditions

Class

Records

Training data

Validation data

PhysioNet/CinC challenge 2017

300

N

N

5154

  

AF

AF

771

  

Others

Non-AF

2557

  

Noisy

 

46

  

Normal

N

918

  

China physiological signal challenge 2018

500

AF (Atrial Fibrillation)

AF

1098

  

I-AVB (First-degree atrioventricular block)

 

704

  

LBBB (Left bundle branch block

 

207

  

RBBB (Right bundle branch block

Non-AF

1695

  

PAC (Premature atrial contraction)

 

574

  

PVC (Premature ventricular contraction)

 

653

  

STD (ST-segment elevated)

 

826

  

STE (ST-segment elevated

 

202

23,714

2635

MIT-BIH trial fibrillation

250

AF (Atrial Fibrillation)

AF

291

  

ECG recording from Chapman University and Shaoxing People’s Hospital

500

SR (Sinus rhythm)

N

1826

  

AF (Atrial fibrillation)

AF

1780

  

SB (Sinus bradycardia)

 

3889

  

ST (Sinus tachycardia)

 

1568

  

AFL (atrial flutter)

 

445

  

SI (Sinus irregularity)

 

399

  

SVT (Supraventricular tachycardia

Non-AF

587

  

AT (Atrial tachycardia)

 

121

  

AVNRT (Atrioventricular node reentrant)

 

16

  

AVRT (Atrioventricular reentrant tachycardia)

 

8

  

SAAWR (Sinus atrium to atrial wandering rhythm)

 

7

  

Indonesia Hospital (ECG 1)

500

Non-AF (other rhythms)

Non-AF

7

 Â