Table 6.

Performance comparison of hybrid CNN-Based models for dental condition detection

Model	Dental Condition	F1-Score	Precision	Recall	Accuracy
CNN + DT	Cavity	0.793 ± 0.028	0.789 ± 0.019	0.798 ± 0.026	0.812 ± 0.029
	Fillings	0.823 ± 0.022	0.834 ± 0.032	0.812 ± 0.024	0.812 ± 0.029
	Impacted Tooth	0.788 ± 0.025	0.784 ± 0.018	0.792 ± 0.026	0.812 ± 0.029
	Implant	0.788 ± 0.030	0.801 ± 0.024	0.776 ± 0.026	0.812 ± 0.029
	Macro Average	0.798 ± 0.026	0.802 ± 0.023	0.795 ± 0.025	0.812 ± 0.029
CNN + RF	Cavity	0.828 ± 0.021	0.823 ± 0.027	0.834 ± 0.023	0.854 ± 0.023
	Fillings	0.860 ± 0.033	0.876 ± 0.022	0.845 ± 0.039	0.854 ± 0.023
	Impacted Tooth	0.854 ± 0.037	0.848 ± 0.029	0.859 ± 0.020	0.854 ± 0.023
	Implant	0.828 ± 0.030	0.839 ± 0.016	0.817 ± 0.037	0.854 ± 0.023
	Macro Average	0.843 ± 0.028	0.847 ± 0.024	0.839 ± 0.030	0.854 ± 0.023
CNN + SVM	Cavity	0.749 ± 0.019	0.745 ± 0.025	0.754 ± 0.032	0.786 ± 0.031
	Fillings	0.787 ± 0.022	0.798 ± 0.024	0.776 ± 0.036	0.786 ± 0.031
	Impacted Tooth	0.759 ± 0.034	0.748 ± 0.034	0.771 ± 0.039	0.786 ± 0.031
	Implant	0.755 ± 0.019	0.762 ± 0.027	0.748 ± 0.023	0.786 ± 0.031
	Macro Average	0.763 ± 0.024	0.763 ± 0.028	0.762 ± 0.033	0.786 ± 0.031