Table 9.
A direct comparison between the existing studies and our findings is based on the performance results, where the short form of ACC, AUC, and TC refers to accuracy, area under the ROC curve, and time complexity, respectively.
| Year and reference | Data collection Source | Number of instances | Type of target class | Reduce imbalance issues | The performing classifiers | Best performingclassifier | The performed results |
|---|---|---|---|---|---|---|---|
|
2022 |
The eICU-CRD (version 2.0) | 2798 | Binary | – |
XGB, LR, RF, SVM |
XGB |
ACC = 82.6%, TC = – |
|
2021 |
Faisalabad Institute of Cardiology | 299 | Binary | SMOTE |
RF, AB, KNN, SVM |
RF |
ACC = 76.25%, TC = – |
|
2021 |
Faisalabad Institute of Cardiology | 299 | Binary | SMOTE | DT, RF, ET, SVM, GB | ET |
ACC = 92.62%, TC = – |
|
2022 |
Faisalabad Institute of Cardiology | 299 | Binary | SMOTE |
SVM, DT, RF |
SVM | ACC = 83.33%, TC = – |
|
2021 |
Ireland and University Hospital of Ioannina | 487 | Multiple | SMOTE | DT, RF, KNN, SVM, LMT, ROT | ROT | ACC = 91.23%, TC = – |
|
2020 |
Faisalabad Institute of Cardiology | 299 | Binary | – |
RF, DT, GB, LR, SVM, KNN, NB |
RF |
ACC = 74%, TC = – |
|
2021 |
The University of California Irvine | 299 | Binary | – |
DT, SVM, KNN, RF |
RF |
ACC = 87%, TC = – |
|
2021 |
Physionet databases |
NA | Multiple | – |
DT, SVM |
SVM | ACC = 88.79%, TC = – |
|
2022 |
Faisalabad Institute of Cardiology | 299 | Binary | SMOTE-ENN | RF, DT, SVM, KNN, LR | RF |
ACC = 90%, TC = – |
|
2023 |
PMRCardio database |
500 | Binary | – |
RF, LR, SVM, GB, XGB |
RF |
ACC = 88%, TC = – |
|
2023 |
Persian Registry Of cardio Vascular diseasE | 2918 | Binary | Undersampling |
DT, RF, XGB LR, SVM, KNN |
XGB |
ACC = 76.4%, TC = – |
|
2022 |
Medical Information Mart for Intensive Care | 46,520 | Binary | – | XGB | XGB |
AUC = 83.1%, TC = – |
|
2019 |
The University of California Irvine | 303 | Binary | – | DT, RF, SVM, GB, HRFLM | HRFLM |
ACC = 88.7%, TC = – |
|
2023 |
Physionet | 2008 | Binary | – | XGB, RF, ET, GB, SVM, KNN, ST | ST |
ACC = 89.41%, TC = – |
|
2019 |
The University of California Irvine | 270 | Binary | – |
LR, NB, MLP, VT |
VT |
ACC = 88.88%, TC = – |
|
2023 Our Study |
Faisalabad Institute of Cardiology | 299 | Binary | BOO-ST |
DT, SVM, ET, KNN, CBCEC |
CBCEC |
ACC = 93.67%, TC = 957 ms |
The signs (–) indicate that the existing studies did not consider specific performance metrics or methods in their model.