IRR-Net: A Joint Learning Framework for Image Reconstruction and Recognition of Photoacoustic Tomography

Table 3

Evaluation metrics for binary classification in the simulation study.

Method		Acc	Sens	Spec	F1 score	Runtime (s)
HH (5 DGD + GoogLeNet)		0.9574	0.9535	0.9691	0.9711	21.4
IRR-Net		0.9474	0.9324	0.9524	0.9572	10.6
HH (DGD + U-Net+ResNet)		0.9474	0.9324	0.9524	0.9572	28.6
LH	1 DGD + GoogLeNet	0.9277	0.9189	0.9486	0.9503	9.3
	TR + GoogLeNet	0.9184	0.9098	0.9440	0.9436	3.2
	TR + AlexNet	0.9012	0.8910	0.9316	0.9311	2.7
	TR + ResNet	0.8987	0.8892	0.9303	0.9275	2.7
HL (5 DGD + SVM)		0.8926	0.8874	0.9271	0.9208	20.1
LL (TR + SVM)		0.8739	0.8665	0.8961	0.9117	1.4