List your company

Lesion Detection on DeepLesion official (test)

76.12Sensitivity @ 0.5 FPPIMULAN
38.36848.16957.9767.771Jun 25, 2018Sep 1, 2018Nov 9, 2018Jan 17, 2019Mar 27, 2019Jun 4, 2019Aug 12, 2019

Evaluation results

MethodLinks
MULANtraining=with 171 tags2019.0876.1283.6988.7692.394.7195.64—85.22
Improved RetinaNet (anchor optimization + dense supervision + attention gate)runtime=41 ms, anchor optimization=true, dense supervision=true, attention gate=true2019.0672.1580.0786.490.7794.0996.32——
RetinaNet + anchor optimization + dense supervisionruntime=39 ms, anchor optimization=true, dense supervision=true2019.0670.2478.2885.190.3993.8196.01——
RetinaNet + anchor optimizationruntime=31 ms, anchor optimization=true2019.0664.8274.9882.2987.8792.294.9——
3DCENumber of slices=272018.0662.4873.3780.785.6589.0991.06——
3DCEruntime=114 ms2019.0662.4873.3780.785.6589.0991.06——
3DCE2019.0862.4873.3780.785.6589.0991.06—75.55
3DCENumber of slices=92018.0659.3270.6879.0984.3487.8189.62——
Data-level fusionNumber of slices=112018.0658.4970.0377.8983.0286.7189.19——
Faster RCNNNumber of slices=32018.0656.967.2675.5781.6285.8388.74——
Faster R-CNNruntime=32 ms2019.0656.967.2675.5781.6285.8388.74——
Improved R-FCNNumber of slices=32018.0656.4967.6576.8982.7687.0389.82——
Original R-FCNNumber of slices=32018.0655.767.2675.3782.2186.2689.19——
ULDor2019.0652.8664.874.8484.3887.1791.8——
ULDor2019.0852.8664.874.8484.3887.1791.8—69.22
No 3D contextNumber of slices=12018.0648.660.5771.1979.1584.7788.42——
original RetinaNetruntime=28 ms2019.0645.854.1762.569.875.3479.48——
Mask R-CNN2019.0639.8252.6665.5877.7385.5491.8——
3DCE2019.08——————75.55—
MULAN2019.08——————86.12—
MULANfeature pyramid=removed2019.08——————79.73—
MULAN3D fusion=removed2019.08——————79.57—
MULANtagging branch=removed2019.08——————84.79—
MULANmask branch=removed2019.08——————85.21—
MULANscore refine. layer=removed2019.08——————84.24—
ULDor2019.08——————69.22—