pISSN 1226-6051
eISSN 2508-786X

Table. 3.

Table. 3.

Summary of AI model performance metrics in cancer study*

Study name AI models Performance metrics

AUC Accuracy Sensitivity Specificity F1 score
Aghakhani et al. (2023) XGBoost 0.83 0.77 0.74 0.77 0.8
Ahamad et al. (2022) RF Medical data
1.00, 0.98a
AE data
1.00+
Medical data
1.00, 0.98a
AE data
1.00+
N/A N/A Medical data
1.00, 0.98a
AE data
1.00+
Banoei et al. (2023) BNN 0.85 N/A 0.57 0.94 N/A
Carbonell et al. (2022) Elastic Net 0.78, 0.82a N/A N/A N/A N/A
An et al. (2020) LASSO 0.83 0.86 0.94 0.90 N/A
Gao et al. (2021) Ensemble model 0.99 0.96 0.87 0.97 0.87
Experton et al. (2021) RF 0.71, 0.66b 0.65, 0.61b N/A N/A N/A
Heydar et al. (2022) RF DM
0.80
non-DM
0.84
DM
0.82
non-DM
0.80
DM
0.80
non-DM
0.91
DM
0.55
non-DM
0.56
N/A
Heyl et al. (2022) RF 0.90 0.83 N/A N/A N/A
Hilal et al. (2022) XGBoost Delta
0.78, 0.81b
Omicron
0.70, 0.78b
Delta
0.96, 0.85b
Omicron
0.98, 0.94b
N/A N/A Delta
0.27, 0.35b:
Omicron
0.27, 0.34b
Ikemura et al. (2021) GBM 0.80 N/A 0.919 0.735 N/A
Jamshidi et al. (2021) RF 0.79 N/A 0.70 0.75 N/A
Edqvist et al. (2023) GBM, RF N/A T1DM
RF: 0.88
T2DM
GBM: 0.74
N/A N/A N/A
Karasneh et al. (2022) LR 0.77 N/A N/A N/A N/A
Lee et al. (2022) LR 0.88 N/A N/A N/A N/A
Modelli de Andrade et al. (2022) Elastic Net 0.78 N/A N/A N/A N/A
Kivrak et al. (2021) XGBoost N/A 0.99 0.99 1.00 N/A
Rahman et al. (2021) self-developed model 0.95 0.90 0.80 0.92 N/A
Lorè et al. (2021) DT 0.73 N/A N/A N/A N/A
Rasmy et al. (2022) self- developed model 0.93, 0.92c N/A N/A N/A N/A
Wollenstein-Betech et al. (2020) LR 0.63, 0.74b 0.79, 0.71b N/A N/A 0.71, 0.70b
Schmidt et al. (2021) XGBoost 0.79 N/A N/A N/A N/A
Alle et al. (2022) LR 0.92 N/A N/A N/A 0.71
Nojiri et al. (2023) Lasso 0.80, 0.78a N/A N/A N/A N/A
Snider et al. (2021) DT 0.93, 0.96a N/A N/A N/A N/A
Kar et al. (2021) XGBoost 0.88 0.97 0.78 0.98 0.81
Wu et al. (2021) self- developed model 0.85 0.75 0.79 0.74 0.40
Guan et al. (2021) XGBoost 1.00 N/A N/A N/A 0.94
Jung et al. (2022) XGBoost 0.65 0.70 N/A N/A N/A
Zhao et al. (2021) SVM 0.94 0.91 0.90 0.94 N/A
Jiao et al. (2021) self- developed model 0•84 N/A 0.73 0.85 0.83
Kang et al. (2021) NN 0.95 N/A 1.00 0.85 0.96
Wong et al. (2021) XGBoost 0.81, 0.72a N/A N/A N/A N/A
Rojas-García et al. (2023) XGBoost 0.79 0.75 0.83 0.74 0.48
Burns et al. (2022) XGBoost 0.75 0.67 N/A 0.66 0.49
Wang et al. (2022) self- developed model 0.85 0.83 0.62 0.89 N/A
Chen et al. (2021) RF 0.90 0.94 0.99 0.93 0.97
De Freitas et al. (2022) RF 0.93 0.90 N/A N/A 0.94
Jehi et al. (2020) self- developed model 0.90 N/A N/A N/A N/A
Hao et al. (2020) RF 0.88b, 0.85c 0.88b, 0.86c N/A N/A 0.91b, 0.91c
Aminu et al. (2022) SVM, LR 1.00 0.99 1.00 0.98 N/A
Chen et al. (2021) LR 0.81 N/A 0.80 0.71 N/A

*all values of predicting mortality except for a: prediction value of severity, b: prediction value of hospitalization, and c: prediction value of mechanical ventilation; +: values including mortality and severity; Lasso: Least Absolute Shrinkage and Selection Operator; XGBoost: Extreme Gradient Boosting; RF: Random Forest; LR: Logistic Regression; DT: Decision Tree; SVM: Support Vector Machine; NN: Neural Network; GBM: Gradient Boosting Machine; AUC: area under the curve; AI: artificial intelligence

Korean J Clin Pharm 2024;34:141-54 https://doi.org/10.24304/kjcp.2024.34.3.141
© 2024 Korean J Clin Pharm