Table 3.

Classification performance of our methods on Tsinghua dataset (%), where 40% AMD and DME volumes are training set, AMD is negative, and DME is positive.

Methods	ACC	SE	SP
FTA-ResNet50 +SVM	$99.93\pm 0.09$	$100.00\pm 0.00$	$99.86\pm 0.18$
FTA-ResNet50 +CNN	$99.49\pm 0.27$	$99.78\pm 0.30$	$99.21\pm 0.42$
FTA-ResNet50+CNN_CBAM	$99.71\pm 0.25$	$99.78\pm 0.30$	$99.64\pm 0.40$