Matches in SemOpenAlex for { <https://semopenalex.org/work/W3159449784> ?p ?o ?g. }
Showing items 1 to 90 of
90
with 100 items per page.
- W3159449784 endingPage "104" @default.
- W3159449784 startingPage "74" @default.
- W3159449784 abstract "Motivation: Traditional image attribution methods struggle to satisfactorily explain predictions of neural networks. Prediction explanation is important, especially in medical imaging, for avoiding the unintended consequences of deploying AI systems when false positive predictions can impact patient care. Thus, there is a pressing need to develop improved models for model explainability and introspection. Specific problem: A new approach is to transform input images to increase or decrease features which cause the prediction. However, current approaches are difficult to implement as they are monolithic or rely on GANs. These hurdles prevent wide adoption. Our approach: Given an arbitrary classifier, we propose a simple autoencoder and gradient update (Latent Shift) that can transform the latent representation of a specific input image to exaggerate or curtail the features used for prediction. We use this method to study chest X-ray classifiers and evaluate their performance. We conduct a reader study with two radiologists assessing 240 chest X-ray predictions to identify which ones are false positives (half are) using traditional attribution maps or our proposed method. Results: We found low overlap with ground truth pathology masks for models with reasonably high accuracy. However, the results from our reader study indicate that these models are generally looking at the correct features. We also found that the Latent Shift explanation allows a user to have more confidence in true positive predictions compared to traditional approaches (0.15$pm$0.95 in a 5 point scale with p=0.01) with only a small increase in false positive predictions (0.04$pm$1.06 with p=0.57). Accompanying webpage: this https URLSource code: this https URL" @default.
- W3159449784 created "2021-05-10" @default.
- W3159449784 creator A5005297246 @default.
- W3159449784 creator A5006028215 @default.
- W3159449784 creator A5035948739 @default.
- W3159449784 creator A5038969629 @default.
- W3159449784 creator A5056076241 @default.
- W3159449784 creator A5060928641 @default.
- W3159449784 creator A5064829377 @default.
- W3159449784 date "2021-02-10" @default.
- W3159449784 modified "2023-09-26" @default.
- W3159449784 title "Gifsplanation via Latent Shift: A Simple Autoencoder Approach to Counterfactual Generation for Chest X-rays" @default.
- W3159449784 hasPublicationYear "2021" @default.
- W3159449784 type Work @default.
- W3159449784 sameAs 3159449784 @default.
- W3159449784 citedByCount "3" @default.
- W3159449784 countsByYear W31594497842021 @default.
- W3159449784 crossrefType "journal-article" @default.
- W3159449784 hasAuthorship W3159449784A5005297246 @default.
- W3159449784 hasAuthorship W3159449784A5006028215 @default.
- W3159449784 hasAuthorship W3159449784A5035948739 @default.
- W3159449784 hasAuthorship W3159449784A5038969629 @default.
- W3159449784 hasAuthorship W3159449784A5056076241 @default.
- W3159449784 hasAuthorship W3159449784A5060928641 @default.
- W3159449784 hasAuthorship W3159449784A5064829377 @default.
- W3159449784 hasConcept C101738243 @default.
- W3159449784 hasConcept C108583219 @default.
- W3159449784 hasConcept C108650721 @default.
- W3159449784 hasConcept C111472728 @default.
- W3159449784 hasConcept C119857082 @default.
- W3159449784 hasConcept C124101348 @default.
- W3159449784 hasConcept C138885662 @default.
- W3159449784 hasConcept C146849305 @default.
- W3159449784 hasConcept C153180895 @default.
- W3159449784 hasConcept C154945302 @default.
- W3159449784 hasConcept C15744967 @default.
- W3159449784 hasConcept C188147891 @default.
- W3159449784 hasConcept C2779458634 @default.
- W3159449784 hasConcept C2780586882 @default.
- W3159449784 hasConcept C41008148 @default.
- W3159449784 hasConcept C64869954 @default.
- W3159449784 hasConcept C77805123 @default.
- W3159449784 hasConcept C95623464 @default.
- W3159449784 hasConceptScore W3159449784C101738243 @default.
- W3159449784 hasConceptScore W3159449784C108583219 @default.
- W3159449784 hasConceptScore W3159449784C108650721 @default.
- W3159449784 hasConceptScore W3159449784C111472728 @default.
- W3159449784 hasConceptScore W3159449784C119857082 @default.
- W3159449784 hasConceptScore W3159449784C124101348 @default.
- W3159449784 hasConceptScore W3159449784C138885662 @default.
- W3159449784 hasConceptScore W3159449784C146849305 @default.
- W3159449784 hasConceptScore W3159449784C153180895 @default.
- W3159449784 hasConceptScore W3159449784C154945302 @default.
- W3159449784 hasConceptScore W3159449784C15744967 @default.
- W3159449784 hasConceptScore W3159449784C188147891 @default.
- W3159449784 hasConceptScore W3159449784C2779458634 @default.
- W3159449784 hasConceptScore W3159449784C2780586882 @default.
- W3159449784 hasConceptScore W3159449784C41008148 @default.
- W3159449784 hasConceptScore W3159449784C64869954 @default.
- W3159449784 hasConceptScore W3159449784C77805123 @default.
- W3159449784 hasConceptScore W3159449784C95623464 @default.
- W3159449784 hasLocation W31594497841 @default.
- W3159449784 hasOpenAccess W3159449784 @default.
- W3159449784 hasPrimaryLocation W31594497841 @default.
- W3159449784 hasRelatedWork W2737498590 @default.
- W3159449784 hasRelatedWork W2891020797 @default.
- W3159449784 hasRelatedWork W2933837888 @default.
- W3159449784 hasRelatedWork W2954283498 @default.
- W3159449784 hasRelatedWork W2962816513 @default.
- W3159449784 hasRelatedWork W2971686358 @default.
- W3159449784 hasRelatedWork W2977389063 @default.
- W3159449784 hasRelatedWork W3102574638 @default.
- W3159449784 hasRelatedWork W3106005960 @default.
- W3159449784 hasRelatedWork W3129345016 @default.
- W3159449784 hasRelatedWork W3130707729 @default.
- W3159449784 hasRelatedWork W3159250167 @default.
- W3159449784 hasRelatedWork W3167454922 @default.
- W3159449784 hasRelatedWork W3172256719 @default.
- W3159449784 hasRelatedWork W3177397812 @default.
- W3159449784 hasRelatedWork W3187853654 @default.
- W3159449784 hasRelatedWork W3191161603 @default.
- W3159449784 hasRelatedWork W3203049632 @default.
- W3159449784 hasRelatedWork W3212900138 @default.
- W3159449784 hasRelatedWork W3213231108 @default.
- W3159449784 isParatext "false" @default.
- W3159449784 isRetracted "false" @default.
- W3159449784 magId "3159449784" @default.
- W3159449784 workType "article" @default.