Skip to main content

Table 4 Detection results of prostate dataset using different methods

From: Enhancing medical image object detection with collaborative multi-agent deep Q-networks and multi-scale representation

Methods

Average IoU (\(\%\))\(\uparrow\)

Wall distance (mm)\(\downarrow\)

Center of mass distance (mm) \(\downarrow\)

Faster R-CNN [35]

80.23

4.63 ± 3.98

4.43 ± 4.86

SSD300 [36]

70

12.24 ± 29.20

14.02 ± 40.86

YOLOv3 [37]

78.34

7.70 ± 21.50

8.95 ± 31.11

DETR [38]

80.36

4.53 ± 4.79

4.41 ± 4.92

DDPG [39]

65.75

9.60 ± 6.90

9.26 ± 6.11

YOLOv4 [40]

80.04

4.75 ± 4.12

4.46 ± 5.58

Our method

80.70

4.39 ± 3.50

4.36 ± 4.15

  1. The bold is used to highlight the best results in the comparisons