Skip to main content
. 2020 May 12;33(4):971–987. doi: 10.1007/s10278-020-00338-w

Table 6.

Retrieval results of Top 10–ranked images with different feature selection methods based on balanced graph similarity

Category AP@10
JMI MRMR CMIM CIFE DISR ICAP
AB 0.2783 0.3349 0.4130 0.3064 0.3847 0.4876
BMP 0.9014 0.8538 0.8442 0.8647 0.8509 0.8573
Calcification 0.9272 0.8720 0.8734 0.8428 0.8638 0.8646
CV 0.9283 0.8948 0.8907 0.8862 0.8862 0.8947
GGO 0.7211 0.6460 0.6372 0.7073 0.6486 0.6309
Lobulation 0.6555 0.6037 0.6257 0.6312 0.6063 0.6421
OP 0.5104 0.4645 0.4471 0.4718 0.4855 0.4870
PI 0.8660 0.8387 0.7887 0.7619 0.7995 0.7980
Speculation 0.5363 0.5015 0.5175 0.5603 0.4803 0.5201
MAP 0.7027 0.6678 0.6706 0.6703 0.6673 0.6869