Skip to main content

Table 5 Comparison with recent studies

From: Developing a high-performance AI model for spontaneous intracerebral hemorrhage mortality prediction using machine learning in ICU settings

Study

This Study

Nie et al., 2021 [21]

Lim et al., 2021 [22]

Guo et al., 2022 [23]

Setting

Intensive care unit

Intensive care unit

In hospital patient

Emergency room

Patient number

1451

760

297

751

Study method

5 ML method

+ 2 scoring system

6 ML method

+ 1 scoring system

2 ML method

+ 1 scoring system

6 ML method

+ 1 scoring system

Feature variables

36 feature variables

72 feature variables

15 feature variables

3–19 feature variables

Outcome

Mortality

Mortality

Mortality

Mortality

Testing results (AUC)

0.913

0.819

0.900

0.844

Best predicting model

XGBoost

Random forest

Support Vector Machine

Logistic regression

Real world application

Yes

No

No

Yes

  1. ML: Machine learning