Table 4 Model performance evaluation.

Feature set	Model name	AUC	Accuracy	Recall	Precision	F1
Radiologic features	Ada boost classifier	0.8517	0.7308	0.8421	0.5926	0.6957
	Logistic regression	0.8469	0.7885	0.6842	0.7222	0.7027
	Random forest	0.8293	0.75	0.4737	0.75	0.5806
	SVM (Radial kernel)	0.8158	0.7692	0.4737	0.8182	0.6
	XGBoost	0.7885	0.8293	0.6316	0.75	0.6857
	KNN	0.7584	0.7115	0.4211	0.6667	0.5161
	Light gradient boosting	0.7257	0.6346	0.3158	0.5	0.3871
	Naive bayes	0.7177	0.7115	0.4737	0.6429	0.5455
Clinical features	Ada boost classifier	0.8222	0.7115	0.5263	0.625	0.5714
	Logistic regression	0.7624	0.7115	0.5263	0.625	0.5714
	Random forest	0.7998	0.6923	0.5263	0.5882	0.5556
	SVM (Radial kernel)	0.7472	0.6923	0.4737	0.6	0.5294
	XGBoost	0.8132	0.75	0.7368	0.6364	0.6829
	KNN	0.8057	0.7115	0.5789	0.6111	0.5946
	Light gradient boosting	0.7616	0.6731	0.5263	0.5556	0.5405
	Naive bayes	0.7352	0.6731	0.6842	0.5417	0.6047
Combination features	Ada boost classifier	0.8947	0.8077	0.7368	0.7368	0.7368
	Logistic regression	0.8628	0.8462	0.6842	0.8667	0.7647
	Random forest	0.8844	0.8462	0.5789	0.9421	0.7333
	SVM (Radial kernel)	0.8612	0.7885	0.6316	0.75	0.6857
	XGBoost	0.8628	0.8077	0.6316	0.8	0.7059
	KNN	0.8158	0.7692	0.6316	0.7059	0.6667
	Light gradient boosting	0.823	0.6923	0.5789	0.5789	0.5789
	Naive bayes	0.8086	0.75	0.7895	0.625	0.6977

Quick links

Search