List your company
ResearchFormerly WizWand

Medical Image Segmentation on MSD Heart (test)

96.85Dice ScoreMetaDCSeg
74.64680.410586.17591.9395Apr 1, 2024Jul 10, 2024Oct 18, 2024Jan 27, 2025May 7, 2025Aug 15, 2025Nov 24, 2025
Updated 5mo ago

Evaluation results

MethodLinks
MetaDCSegNoise Level=0%2025.1196.85—93.892.12
MetaDCSegNoise Level=20%2025.1195.89—92.152.89
L2B*Noise Level=0%2025.1195.42—91.233.45
MetaDCSegNoise Level=40%2025.1194.87—90.343.78
L2B*Noise Level=20%2025.1194.38—89.454.23
MADGNet*Noise Level=0%2025.1193.68—88.124.89
MAXFormer*Noise Level=0%2025.1193.24—87.345.12
ConDSeg*Noise Level=0%2025.1193.06—87.055.45
HAMIL*Noise Level=0%2025.1192.98—86.925.67
L2B*Noise Level=40%2025.1192.89—86.785.67
MetaDCSegNoise Level=60%2025.1192.76—86.565.12
MADGNet*Noise Level=20%2025.1192.08—85.346.78
MAXFormer*Noise Level=20%2025.1191.35—84.157.45
ConDSeg*Noise Level=20%2025.1190.89—83.568.12
HAMIL*Noise Level=20%2025.1190.12—82.458.56
L2B*Noise Level=60%2025.1189.62—81.238.45
MADGNet*Noise Level=40%2025.1189.15—80.459.45
MAXFormer*Noise Level=40%2025.1188.02—78.6210.34
ConDSeg*Noise Level=40%2025.1187.54—77.8911.23
HAMIL*Noise Level=40%2025.1186.56—76.3412.45
MADGNet*Noise Level=60%2025.1183.78—72.1214.56
UVI-Net (w/o inst opt.)Augmentation Scale=10x, Aggregation=Average of 3D-UNet, VNet, and UNETR, Note=w/o instance optimization2024.0483.1———
ConDSeg*Noise Level=60%2025.1182.15—69.7815.89
DDMAugmentation Scale=10x, Aggregation=Average of 3D-UNet, VNet, and UNETR2024.0481.8———
MAXFormer*Noise Level=60%2025.1181.25—68.4516.78
Fourier-Net+Augmentation Scale=10x, Aggregation=Average of 3D-UNet, VNet, and UNETR2024.0480.9———
VMAugmentation Scale=10x, Aggregation=Average of 3D-UNet, VNet, and UNETR2024.0480.3———
HAMIL*Noise Level=60%2025.1180.14—66.8918.23
R2NetAugmentation Scale=10x, Aggregation=Average of 3D-UNet, VNet, and UNETR2024.0478.9———
TMAugmentation Scale=10x, Aggregation=Average of 3D-UNet, VNet, and UNETR2024.0477.3———
VanillaAugmentation Scale=10x, Aggregation=Average of 3D-UNet, VNet, and UNETR2024.0475.5———
nnU-NetEvaluation Mode=Test Set Submission2018.09—92.77——