Table 7.

Comparison of our model with models in previous studies.

Study (data source)	Number of patients	Accuracy (%)	AUCa (%)	MAEb (%)	Precision (%)	Sensitivity (%)	Specificity (%)	F1-score (%)
Brahim et al [1] (images)	1024	0.83	—c	—	—	0.87	0.81	—
Shen et al [2] (stats)	195	—	—	0.11	—	0.71	0.67	0.69
Halilaj et al [19] (stats)	3924	—	0.95	—	—	—	—	—
Yong et al [17] (images)	4796	0.88	—	0.33	—	—	—	—
Lim et al [6] (stats)	5749	0.72	0.77	—	—	0.67	0.73	—
Tiulpin et al [22] (images)	4928	—	0.79	—	—	—	—	—
Our study	5849	0.91	0.91	0.09	0.91	0.94	0.87	0.93

^aAUC: area under the curve.

^bMAE: mean absolute error.

^cNot applicable.