Thyroid Diagnosis from SPECT Images Using Convolutional Neural Network with Optimization

Table 1

Grave’s disease class performance comparison of different methods (percent).

Network	DenseNet121	ResNet101	InceptionV3	VGG19	MVGG	GoogleNet	SDAE	Proposed
Recall	95.17	93.83	88.50	89.00	89.50	90.83	92.33	97.50
Precision	98.11	98.57	90.15	91.44	91.95	91.91	94.22	98.82
Accuracy	98.33	97.63	94.71	95.17	95.42	95.71	96.67	99.08
Specificity	99.39	98.89	96.78	97.22	97.39	97.33	98.11	99.61
F1 score	96.62	95.18	89.32	90.20	90.71	91.37	93.27	98.15