Table 2. MAE, RMSE, and t-test results for each model, comparing the actual bone age to the predicted bone age for both the original images and the images subjected to the contrast conversion algorithm

Model	Items	Original	Image contrast conversion methods
			CLAHE	FCE	HE
CNN	MAE	26.05	25.81	32.21	31.74
	RMSE	2.60	2.58	3.22	3.17
	P-value		<0.05	0.22	<0.05
ResNet 50	MAE	43.69	46.58	44.63	54.35
	RMSE	4.37	4.66	4.46	5.43
	P-value		<0.05	<0.05	<0.05
VGG 19	MAE	34.26	36.09	34.86	33.08
	RMSE	3.43	3.61	3.49	3.31
	P-value		<0.05	<0.05	<0.05
Inception V3	MAE	34.54	36.03	34.94	33.29
	RMSE	3.45	3.60	3.49	3.33
	P-value		0.25	<0.05	<0.05
Xception	MAE	32.24	30.13	31.64	34.13
	RMSE	3.22	3.01	3.06	3.41
	P-value		0.49	0.19	0.05

MAE, mean absolute error; RMSE, root mean square error; CLAHE, contrast limited adaptive histogram equalization; FCE, fuzzy contrast enhancement; HE, histogram equalization; CNN, convolutional neural network; ResNet 50, Residual Network 50; VGG 19, Visual Geometry Group 19.