TABLE IV.
The Effect of Bigaug With Big Data (465 MRI Volumes From Multiple Medial Centers, MRI Vendors, and Protocols) for the Task of Whole Prostate Segmentation in MRI Volumes. Note That the State-of-the-Art Methods Marked With
Source | Unseen | ||||||
---|---|---|---|---|---|---|---|
train | val | MSD-P | PROMISE | NCI-ISBI | ProstateX | Average | |
Baseline | 95.6 | 89.9 | 87.8 | 82.9 | 88.8 | 90.6 | 87.5 |
BigAug (ours) | 94.1 | 91.8 | 89.1 | 88.1 | 89.4 | 91.9 | 89.6 |
State-of-the-art | - | - | - | 91.4* [39] | 89.3* [40] | 91.9* | 90.9 |
are Trained And Tested on the Same Domain or Inter-Observer Variability (91.9%). No Evaluation of the Whole Prostate Segmentation is Available in MSD Challenge (http://medicaldecathlon.com/results.html)