Skip to main content

Table 4 Classification of the features of the included articles

From: Cervical cancer survival prediction by machine learning algorithms: a systematic review

Characteristics

Categories

Number (n)

OS

DFS

PFS

Location

Asia

5 [17, 19,20,21, 23]

5 [16, 18, 21,22,23]

1 [19]

Europe

2 [26, 27]

1 [24]

1 [25]

USA

1 [28]

-

1 [28]

Dataset sources

Hospitals

6 [19, 21, 23, 26,27,28]

5 [16, 21,22,23,24]

3 [19, 25, 28]

SEER

1 [17]

-

-

TCGA

1 [20]

-

-

GEO

-

1 [18]

-

Dataset privacy

Public

2 [17, 20]

1 [18]

-

Private

6 [19, 21, 23, 26,27,28]

5 [16, 21,22,23,24]

3 [19, 25, 28]

Data source

Single

6 [17, 20, 23, 26,27,28]

5 [16, 18, 22,23,24]

2 [25, 28]

Multiple

2 [19, 21]

1 [21]

1 [19]

Preprocessing

Yes

6 [17, 19, 20, 23, 26, 27]

5 [16, 18, 22,23,24]

1 [19]

No

2 [21, 28]

1 [21]

2 [25, 28]

Feature selection

Yes

7 [17, 19,20,21, 23, 26, 28]

6 [16, 18, 21,22,23,24]

3 [19, 25, 28]

No

1 [27]

-

-

# Models

One

4 [17, 20, 23, 26]

3 [16, 22, 23]

-

Two or more

4 [19, 21, 27, 28]

3 [18, 21, 24]

3 [19, 25, 28]

Models type

RF

3 [21, 26, 28]

3 [16, 21, 24]

2 [25, 28]

LR

1 [19]

2 [17, 24]

2 [19, 25]

SVM

2 [20, 27]

1 [24]

-

DL

2 [23, 28]

2 [22, 23]

-

H&E L

2 [19, 21]

2 [18, 21]

1 [19]

Validation

Internal

8 [17, 19,20,21, 23, 26,27,28]

4 [16, 21,22,23]

3 [19, 25, 28]

External

-

2 [18, 24]

-

Evaluation metrics

AUC

4 [19, 20, 23, 27]

4 [16, 19, 23, 24]

1 [25]

C-index

5 [17, 19, 21, 23, 26]

2 [21, 23]

1 [19]

Sensitivity

1 [27]

3 [18, 22, 24]

-

Precision

-

2 [18, 24]

1 [25]

Specificity

1 [27]

1 [22]

-

Accuracy

1 [27]

2 [18, 22]

1 [25]

F1-score

-

2 [18, 24]

-

MAE

2 [21, 28]

1 [21]

1 [28]

NPV / PPV

-

1 [22]

-

Data types

Clinical

5 [17, 19, 21, 27, 28]

1 [21]

2 [19, 28]

Image

1 [26]

-

-

Molecular

-

1 [18]

-

Clinical + Image

1 [23]

4 [16, 22,23,24]

1 [25]

Clinical + Molecular

1 [20]

-

-

  1. TCGA The cancer Genome Atlas, GEO Gene Expression Omnibus, SEER Surveillance, Epidemiology, and End Results, RF Random Forest, SVM Support Vector Machine, LR Logistic regression, DL Deep Learning, H&E L Hybrid and Ensemble learning, MAE Mean absolute error