Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386799758> ?p ?o ?g. }
- W4386799758 endingPage "102967" @default.
- W4386799758 startingPage "102967" @default.
- W4386799758 abstract "Any clinically-deployed image-processing pipeline must be robust to the full range of inputs it may be presented with. One popular approach to this challenge is to develop predictive models that can provide a measure of their uncertainty. Another approach is to use generative modelling to quantify the likelihood of inputs. Inputs with a low enough likelihood are deemed to be out-of-distribution and are not presented to the downstream predictive model. In this work, we evaluate several approaches to segmentation with uncertainty for the task of segmenting bleeds in 3D CT of the head. We show that these models can fail catastrophically when operating in the far out-of-distribution domain, often providing predictions that are both highly confident and wrong. We propose to instead perform out-of-distribution detection using the Latent Transformer Model: a VQ-GAN is used to provide a highly compressed latent representation of the input volume, and a transformer is then used to estimate the likelihood of this compressed representation of the input. We demonstrate this approach can identify images that are both far- and near- out-of-distribution, as well as provide spatial maps that highlight the regions considered to be out-of-distribution. Furthermore, we find a strong relationship between an image's likelihood and the quality of a model's segmentation on it, demonstrating that this approach is viable for filtering out unsuitable images." @default.
- W4386799758 created "2023-09-17" @default.
- W4386799758 creator A5014517800 @default.
- W4386799758 creator A5023962528 @default.
- W4386799758 creator A5029342491 @default.
- W4386799758 creator A5035697250 @default.
- W4386799758 creator A5047180446 @default.
- W4386799758 creator A5050542038 @default.
- W4386799758 creator A5051886778 @default.
- W4386799758 creator A5055334565 @default.
- W4386799758 creator A5057221976 @default.
- W4386799758 creator A5074184782 @default.
- W4386799758 creator A5074881456 @default.
- W4386799758 creator A5077080413 @default.
- W4386799758 creator A5082106258 @default.
- W4386799758 creator A5083427047 @default.
- W4386799758 creator A5088937468 @default.
- W4386799758 date "2023-12-01" @default.
- W4386799758 modified "2023-10-03" @default.
- W4386799758 title "Latent Transformer Models for out-of-distribution detection" @default.
- W4386799758 cites W2036200541 @default.
- W4386799758 cites W2112796928 @default.
- W4386799758 cites W2804939189 @default.
- W4386799758 cites W2900936384 @default.
- W4386799758 cites W2962785568 @default.
- W4386799758 cites W2971561076 @default.
- W4386799758 cites W2980702656 @default.
- W4386799758 cites W3098712157 @default.
- W4386799758 cites W3153261712 @default.
- W4386799758 cites W3172681723 @default.
- W4386799758 cites W3172942063 @default.
- W4386799758 cites W3180355996 @default.
- W4386799758 cites W4225370003 @default.
- W4386799758 cites W4312933868 @default.
- W4386799758 cites W4385801725 @default.
- W4386799758 doi "https://doi.org/10.1016/j.media.2023.102967" @default.
- W4386799758 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/37778102" @default.
- W4386799758 hasPublicationYear "2023" @default.
- W4386799758 type Work @default.
- W4386799758 citedByCount "0" @default.
- W4386799758 crossrefType "journal-article" @default.
- W4386799758 hasAuthorship W4386799758A5014517800 @default.
- W4386799758 hasAuthorship W4386799758A5023962528 @default.
- W4386799758 hasAuthorship W4386799758A5029342491 @default.
- W4386799758 hasAuthorship W4386799758A5035697250 @default.
- W4386799758 hasAuthorship W4386799758A5047180446 @default.
- W4386799758 hasAuthorship W4386799758A5050542038 @default.
- W4386799758 hasAuthorship W4386799758A5051886778 @default.
- W4386799758 hasAuthorship W4386799758A5055334565 @default.
- W4386799758 hasAuthorship W4386799758A5057221976 @default.
- W4386799758 hasAuthorship W4386799758A5074184782 @default.
- W4386799758 hasAuthorship W4386799758A5074881456 @default.
- W4386799758 hasAuthorship W4386799758A5077080413 @default.
- W4386799758 hasAuthorship W4386799758A5082106258 @default.
- W4386799758 hasAuthorship W4386799758A5083427047 @default.
- W4386799758 hasAuthorship W4386799758A5088937468 @default.
- W4386799758 hasBestOaLocation W43867997581 @default.
- W4386799758 hasConcept C107673813 @default.
- W4386799758 hasConcept C119857082 @default.
- W4386799758 hasConcept C121332964 @default.
- W4386799758 hasConcept C153180895 @default.
- W4386799758 hasConcept C154945302 @default.
- W4386799758 hasConcept C165801399 @default.
- W4386799758 hasConcept C167966045 @default.
- W4386799758 hasConcept C177769412 @default.
- W4386799758 hasConcept C39890363 @default.
- W4386799758 hasConcept C41008148 @default.
- W4386799758 hasConcept C62520636 @default.
- W4386799758 hasConcept C66322947 @default.
- W4386799758 hasConcept C89600930 @default.
- W4386799758 hasConceptScore W4386799758C107673813 @default.
- W4386799758 hasConceptScore W4386799758C119857082 @default.
- W4386799758 hasConceptScore W4386799758C121332964 @default.
- W4386799758 hasConceptScore W4386799758C153180895 @default.
- W4386799758 hasConceptScore W4386799758C154945302 @default.
- W4386799758 hasConceptScore W4386799758C165801399 @default.
- W4386799758 hasConceptScore W4386799758C167966045 @default.
- W4386799758 hasConceptScore W4386799758C177769412 @default.
- W4386799758 hasConceptScore W4386799758C39890363 @default.
- W4386799758 hasConceptScore W4386799758C41008148 @default.
- W4386799758 hasConceptScore W4386799758C62520636 @default.
- W4386799758 hasConceptScore W4386799758C66322947 @default.
- W4386799758 hasConceptScore W4386799758C89600930 @default.
- W4386799758 hasLocation W43867997581 @default.
- W4386799758 hasLocation W43867997582 @default.
- W4386799758 hasOpenAccess W4386799758 @default.
- W4386799758 hasPrimaryLocation W43867997581 @default.
- W4386799758 hasRelatedWork W2064275965 @default.
- W4386799758 hasRelatedWork W2138073421 @default.
- W4386799758 hasRelatedWork W2961085424 @default.
- W4386799758 hasRelatedWork W3128025265 @default.
- W4386799758 hasRelatedWork W4285260836 @default.
- W4386799758 hasRelatedWork W4286629047 @default.
- W4386799758 hasRelatedWork W4306321456 @default.
- W4386799758 hasRelatedWork W4306674287 @default.
- W4386799758 hasRelatedWork W4308847496 @default.
- W4386799758 hasRelatedWork W4224009465 @default.
- W4386799758 hasVolume "90" @default.