Hybrid deep learning and graph matching segments whole hearts and great vessels in congenital heart disease CT images, achieving 11.9% higher average Dice score than prior methods on 68 scans across 14 CHD types.
Scaling for edge inference of deep neural networks,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.IV 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Accurate Congenital Heart Disease Model Generation for 3D Printing
Hybrid deep learning and graph matching segments whole hearts and great vessels in congenital heart disease CT images, achieving 11.9% higher average Dice score than prior methods on 68 scans across 14 CHD types.