Skip to main content

Table 2 The configuration of the feature vector

From: An arrhythmia classification algorithm using a dedicated wavelet adapted to different subjects

No

Feature (abbreviation)

Dimension

Description

Equation

1

RRI[k]

4

Adjacent RR intervals

k = -2 t(R peak [n - 2]) - t(R peak [n - 3])

    

k = -1 t(R peak [n - 1]) - t(R peak [n - 2])

    

k = 0 t(R peak [n]) - t(R peak [n - 1])

    

k = 1 t(R peak [n + 1]) - t(R peak [n])

2

RRIavg

1

The average RRI of the last 10 heartbeats

3

RRIdiv

1

The standard deviation between the RRI of last 10 heartbeats

4

Annotbf1

5

The classification results of the last heartbeat

[1 0 0 0 0] if Y n = Normal beat

    

[0 0 1 0 0] if Y n = SVEB

    

[0 0 1 0 0] if Y n = VEB

    

[0 0 0 1 0] if Y n = Fusion beat

    

[0 0 0 0 1] if Y n = Unknown beat

5

Annotbf2

5

The classification results of the second-to-last heartbeat

 

6

Morph

4

Morphology data of the current heartbeat (After CWT-PCA-LDA)

LDA(PCA(CWT(X n )))

7

Morphbf1

4

Morphology data of the last heartbeat (After CWT-PCA-LDA)

LDA(PCA(CWT(X n-1)))

8

Morphbf2

4

Morphology data of the second-to-last heartbeat (After CWT-PCA-LDA)

LDA(PCA(CWT(X n-2)))