Skip to main content

Table 2 Summary of the most related works and current work

From: Deep learning based fetal distress detection from time frequency representation of cardiotocogram signal using Morse wavelet: research study

Paper

Database Source

Labeling criteria

Stage

Method

Performance

Bursa et al. [16]

CTU-UHB

pH

1st (60 min)

CWT + 2D CNN

Acc: 94.1%

Comert et al. [17]

CTU-UHB

pH

2nd (15 min)

STFT + 2D CNN (AlexNet)

Acc: 93.32%, Sen: 56.15% Sp: 96.51%

Zhao et al. [18]

CTU-UHB

pH

1st (20 min)

CWT + 2D CNN

Acc: 98.34%, Sen: 94.87% Sp: 97.82%

Parvathavarthine et al. [19]

CTU-UHB

pH

2nd (30 min)

Optimized Res Net 50

Acc: 94.63%

Frasch et al. [20]

Private

Visual

N/A

Customized 2D CNN

Acc: 93.6%

Ogasawara et al. [21]

CTU-UHB

Ph&apgar

2nd (30 min)

2D CTG-net

AUC: 0.73

Saini et al. [22]

CTU-UHB

pH

1st (60 min)

FHR + UC + 2D CNN

Acc: 70%, 71.4% and 70%

Current study

CTU-UHB

pH

1st (20 min) and 2nd (15 min)

Morse wavelet + CNN(ResNet50)

Acc: 98.7%,

Sen.: 97.0%

Sp: 100%

     

Acc: 96.1%,

Sen.: 94.1%

Sp: 97.7%