Matches in SemOpenAlex for { <https://semopenalex.org/work/W3204235424> ?p ?o ?g. }
- W3204235424 endingPage "314" @default.
- W3204235424 startingPage "304" @default.
- W3204235424 abstract "Despite their outstanding accuracy, semi-supervised segmentation methods based on deep neural networks can still yield predictions that are considered anatomically impossible by clinicians, for instance, containing holes or disconnected regions. To solve this problem, we present a Context-aware Virtual Adversarial Training (CaVAT) method for generating anatomically plausible segmentation. Unlike approaches focusing solely on accuracy, our method also considers complex topological constraints like connectivity which cannot be easily modeled in a differentiable loss function. We use adversarial training to generate examples violating the constraints, so the network can learn to avoid making such incorrect predictions on new examples, and employ the Reinforce algorithm to handle non-differentiable segmentation constraints. The proposed method offers a generic and efficient way to add any constraint on top of any segmentation network. Experiments on two clinically-relevant datasets show our method to produce segmentations that are both accurate and anatomically-plausible in terms of region connectivity." @default.
- W3204235424 created "2021-10-11" @default.
- W3204235424 creator A5013736647 @default.
- W3204235424 creator A5017608284 @default.
- W3204235424 creator A5038088018 @default.
- W3204235424 creator A5039660800 @default.
- W3204235424 creator A5051873577 @default.
- W3204235424 creator A5053557557 @default.
- W3204235424 date "2021-01-01" @default.
- W3204235424 modified "2023-10-06" @default.
- W3204235424 title "Context-Aware Virtual Adversarial Training for Anatomically-Plausible Segmentation" @default.
- W3204235424 cites W1783315696 @default.
- W3204235424 cites W2106033751 @default.
- W3204235424 cites W2620296437 @default.
- W3204235424 cites W2750925197 @default.
- W3204235424 cites W2778764040 @default.
- W3204235424 cites W2799738340 @default.
- W3204235424 cites W2804047627 @default.
- W3204235424 cites W2888570268 @default.
- W3204235424 cites W2914410118 @default.
- W3204235424 cites W2918552801 @default.
- W3204235424 cites W2921629474 @default.
- W3204235424 cites W2962736273 @default.
- W3204235424 cites W2963073217 @default.
- W3204235424 cites W2963785012 @default.
- W3204235424 cites W2964159205 @default.
- W3204235424 cites W2964250774 @default.
- W3204235424 cites W2979433110 @default.
- W3204235424 cites W2982043723 @default.
- W3204235424 cites W3000109616 @default.
- W3204235424 cites W3005650525 @default.
- W3204235424 cites W3032691033 @default.
- W3204235424 cites W3042302007 @default.
- W3204235424 cites W3101848196 @default.
- W3204235424 cites W3102538446 @default.
- W3204235424 cites W3106105822 @default.
- W3204235424 doi "https://doi.org/10.1007/978-3-030-87193-2_29" @default.
- W3204235424 hasPublicationYear "2021" @default.
- W3204235424 type Work @default.
- W3204235424 sameAs 3204235424 @default.
- W3204235424 citedByCount "4" @default.
- W3204235424 countsByYear W32042354242022 @default.
- W3204235424 countsByYear W32042354242023 @default.
- W3204235424 crossrefType "book-chapter" @default.
- W3204235424 hasAuthorship W3204235424A5013736647 @default.
- W3204235424 hasAuthorship W3204235424A5017608284 @default.
- W3204235424 hasAuthorship W3204235424A5038088018 @default.
- W3204235424 hasAuthorship W3204235424A5039660800 @default.
- W3204235424 hasAuthorship W3204235424A5051873577 @default.
- W3204235424 hasAuthorship W3204235424A5053557557 @default.
- W3204235424 hasBestOaLocation W32042354242 @default.
- W3204235424 hasConcept C119857082 @default.
- W3204235424 hasConcept C134306372 @default.
- W3204235424 hasConcept C14036430 @default.
- W3204235424 hasConcept C151730666 @default.
- W3204235424 hasConcept C153180895 @default.
- W3204235424 hasConcept C154945302 @default.
- W3204235424 hasConcept C202615002 @default.
- W3204235424 hasConcept C2524010 @default.
- W3204235424 hasConcept C2776036281 @default.
- W3204235424 hasConcept C2779343474 @default.
- W3204235424 hasConcept C2984842247 @default.
- W3204235424 hasConcept C33923547 @default.
- W3204235424 hasConcept C37736160 @default.
- W3204235424 hasConcept C41008148 @default.
- W3204235424 hasConcept C50644808 @default.
- W3204235424 hasConcept C78458016 @default.
- W3204235424 hasConcept C86803240 @default.
- W3204235424 hasConcept C89600930 @default.
- W3204235424 hasConceptScore W3204235424C119857082 @default.
- W3204235424 hasConceptScore W3204235424C134306372 @default.
- W3204235424 hasConceptScore W3204235424C14036430 @default.
- W3204235424 hasConceptScore W3204235424C151730666 @default.
- W3204235424 hasConceptScore W3204235424C153180895 @default.
- W3204235424 hasConceptScore W3204235424C154945302 @default.
- W3204235424 hasConceptScore W3204235424C202615002 @default.
- W3204235424 hasConceptScore W3204235424C2524010 @default.
- W3204235424 hasConceptScore W3204235424C2776036281 @default.
- W3204235424 hasConceptScore W3204235424C2779343474 @default.
- W3204235424 hasConceptScore W3204235424C2984842247 @default.
- W3204235424 hasConceptScore W3204235424C33923547 @default.
- W3204235424 hasConceptScore W3204235424C37736160 @default.
- W3204235424 hasConceptScore W3204235424C41008148 @default.
- W3204235424 hasConceptScore W3204235424C50644808 @default.
- W3204235424 hasConceptScore W3204235424C78458016 @default.
- W3204235424 hasConceptScore W3204235424C86803240 @default.
- W3204235424 hasConceptScore W3204235424C89600930 @default.
- W3204235424 hasLocation W32042354241 @default.
- W3204235424 hasLocation W32042354242 @default.
- W3204235424 hasOpenAccess W3204235424 @default.
- W3204235424 hasPrimaryLocation W32042354241 @default.
- W3204235424 hasRelatedWork W2755252132 @default.
- W3204235424 hasRelatedWork W2914158293 @default.
- W3204235424 hasRelatedWork W3046843850 @default.
- W3204235424 hasRelatedWork W3182705599 @default.
- W3204235424 hasRelatedWork W4288261383 @default.
- W3204235424 hasRelatedWork W4289711245 @default.
- W3204235424 hasRelatedWork W4311734044 @default.