Can artificial intelligence detect type 2 diabetes in women by evaluating the pectoral muscle on tomosynthesis: diagnostic study

Insights into Imaging

Table 2 Accuracy and loss values of the training iterations

*Epoch 15 steps per epoch = 1000* *Model/data fit generator* *iphyton input*	*ms/step*	*loss*	*Accuracy*	*Validation data loss*	*Validation accuracy*	*callbacks*
Epoch 1/15 - 167s	141	1.1041	0.3590	1.0632	0.4362	5.000e−05
Epoch 2/15 - 135s	135	1.0670	0.4340	1.0191	0.5046	5.000e−05
Epoch 3/15 - 136s	136	1.0043	0.5110	0.9447	0.5552	5.000e−05
Epoch 4/15 - 136s	136	0.9489	0.5495	0.8263	0.6282	5.000e−05
Epoch 5/15 - 136s	136	0.8795	0.6005	0.8034	0.6448	5.000e−05
Epoch 6/15 - 136s	136	0.8193	0.6320	0.6383	0.7305	5.000e−05
Epoch 7/15 - 136s	136	0.7507	0.6810	0.5571	0.7856	5.000e−05
Epoch 8/15 - 136s	136	0.6613	0.7220	0.4912	0.7897	5.000e−05
Epoch 9/15 - 136s	136	0.5542	0.7660	0.4433	0.8103	5.000e−05
Epoch 10/15 - 136s	136	0.5055	0.7975	0.3150	0.8885	5.000e−05
Epoch 11/15 - 136s	136	0.3943	0.8500	0.2988	0.8931	5.000e−05
Epoch 12/15 - 136s	136	0.3728	0.8570	0.2376	0.9224	5.000e−05
Epoch 13/15 - 136s	136	0.3124	0.8800	0.2040	0.9322	5.000e−05
Epoch 14/15 - 136s	136	0.2849	0.8925	0.1816	0.9368	5.000e−05
Epoch 15/15 - 137s	137	0.2202	0.9245	0.1427	0.9454	5.000e−05