FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W3003770199> ?p ?o ?g. }

• W3003770199 endingPage "101660" @default.
• W3003770199 startingPage "101660" @default.
• W3003770199 abstract "In the last decades, large datasets of fundus photographs have been collected in diabetic retinopathy (DR) screening networks. Through deep learning, these datasets were used to train automatic detectors for DR and a few other frequent pathologies, with the goal to automate screening. One challenge limits the adoption of such systems so far: automatic detectors ignore rare conditions that ophthalmologists currently detect, such as papilledema or anterior ischemic optic neuropathy. The reason is that standard deep learning requires too many examples of these conditions. However, this limitation can be addressed with few-shot learning, a machine learning paradigm where a classifier has to generalize to a new category not seen in training, given only a few examples of this category. This paper presents a new few-shot learning framework that extends convolutional neural networks (CNNs), trained for frequent conditions, with an unsupervised probabilistic model for rare condition detection. It is based on the observation that CNNs often perceive photographs containing the same anomalies as similar, even though these CNNs were trained to detect unrelated conditions. This observation was based on the t-SNE visualization tool, which we decided to incorporate in our probabilistic model. Experiments on a dataset of 164,660 screening examinations from the OPHDIAT screening network show that 37 conditions, out of 41, can be detected with an area under the ROC curve (AUC) greater than 0.8 (average AUC: 0.938). In particular, this framework significantly outperforms other frameworks for detecting rare conditions, including multitask learning, transfer learning and Siamese networks, another few-shot learning solution. We expect these richer predictions to trigger the adoption of automated eye pathology screening, which will revolutionize clinical practice in ophthalmology." @default.
• W3003770199 created "2020-02-07" @default.
• W3003770199 creator A5028479581 @default.
• W3003770199 creator A5031442831 @default.
• W3003770199 creator A5040529738 @default.
• W3003770199 creator A5052923096 @default.
• W3003770199 creator A5089956527 @default.
• W3003770199 date "2020-04-01" @default.
• W3003770199 modified "2023-10-18" @default.
• W3003770199 title "Automatic detection of rare pathologies in fundus photographs using few-shot learning" @default.
• W3003770199 cites W2035318430 @default.
• W3003770199 cites W2114605070 @default.
• W3003770199 cites W2115531174 @default.
• W3003770199 cites W2115733720 @default.
• W3003770199 cites W2118020555 @default.
• W3003770199 cites W2123215356 @default.
• W3003770199 cites W2167655509 @default.
• W3003770199 cites W2529153069 @default.
• W3003770199 cites W2557738935 @default.
• W3003770199 cites W2592929672 @default.
• W3003770199 cites W2598442119 @default.
• W3003770199 cites W2767040741 @default.
• W3003770199 cites W2772246530 @default.
• W3003770199 cites W2792026451 @default.
• W3003770199 cites W2886281300 @default.
• W3003770199 cites W2888933418 @default.
• W3003770199 cites W2893356526 @default.
• W3003770199 cites W2896966666 @default.
• W3003770199 cites W2899170201 @default.
• W3003770199 cites W2899951262 @default.
• W3003770199 cites W2912072952 @default.
• W3003770199 cites W2913340405 @default.
• W3003770199 cites W2917093036 @default.
• W3003770199 cites W2938600289 @default.
• W3003770199 cites W2952436003 @default.
• W3003770199 cites W4247066598 @default.
• W3003770199 doi "https://doi.org/10.1016/j.media.2020.101660" @default.
• W3003770199 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/32028213" @default.
• W3003770199 hasPublicationYear "2020" @default.
• W3003770199 type Work @default.
• W3003770199 sameAs 3003770199 @default.
• W3003770199 citedByCount "49" @default.
• W3003770199 countsByYear W30037701992020 @default.
• W3003770199 countsByYear W30037701992021 @default.
• W3003770199 countsByYear W30037701992022 @default.
• W3003770199 countsByYear W30037701992023 @default.
• W3003770199 crossrefType "journal-article" @default.
• W3003770199 hasAuthorship W3003770199A5028479581 @default.
• W3003770199 hasAuthorship W3003770199A5031442831 @default.
• W3003770199 hasAuthorship W3003770199A5040529738 @default.
• W3003770199 hasAuthorship W3003770199A5052923096 @default.
• W3003770199 hasAuthorship W3003770199A5089956527 @default.
• W3003770199 hasBestOaLocation W30037701991 @default.
• W3003770199 hasConcept C108583219 @default.
• W3003770199 hasConcept C118487528 @default.
• W3003770199 hasConcept C119857082 @default.
• W3003770199 hasConcept C12267149 @default.
• W3003770199 hasConcept C150899416 @default.
• W3003770199 hasConcept C153180895 @default.
• W3003770199 hasConcept C154945302 @default.
• W3003770199 hasConcept C189119545 @default.
• W3003770199 hasConcept C2776391266 @default.
• W3003770199 hasConcept C36464697 @default.
• W3003770199 hasConcept C41008148 @default.
• W3003770199 hasConcept C49937458 @default.
• W3003770199 hasConcept C52001869 @default.
• W3003770199 hasConcept C71924100 @default.
• W3003770199 hasConcept C81363708 @default.
• W3003770199 hasConcept C95623464 @default.
• W3003770199 hasConceptScore W3003770199C108583219 @default.
• W3003770199 hasConceptScore W3003770199C118487528 @default.
• W3003770199 hasConceptScore W3003770199C119857082 @default.
• W3003770199 hasConceptScore W3003770199C12267149 @default.
• W3003770199 hasConceptScore W3003770199C150899416 @default.
• W3003770199 hasConceptScore W3003770199C153180895 @default.
• W3003770199 hasConceptScore W3003770199C154945302 @default.
• W3003770199 hasConceptScore W3003770199C189119545 @default.
• W3003770199 hasConceptScore W3003770199C2776391266 @default.
• W3003770199 hasConceptScore W3003770199C36464697 @default.
• W3003770199 hasConceptScore W3003770199C41008148 @default.
• W3003770199 hasConceptScore W3003770199C49937458 @default.
• W3003770199 hasConceptScore W3003770199C52001869 @default.
• W3003770199 hasConceptScore W3003770199C71924100 @default.
• W3003770199 hasConceptScore W3003770199C81363708 @default.
• W3003770199 hasConceptScore W3003770199C95623464 @default.
• W3003770199 hasLocation W30037701991 @default.
• W3003770199 hasLocation W30037701992 @default.
• W3003770199 hasLocation W30037701993 @default.
• W3003770199 hasLocation W30037701994 @default.
• W3003770199 hasLocation W30037701995 @default.
• W3003770199 hasLocation W30037701996 @default.
• W3003770199 hasOpenAccess W3003770199 @default.
• W3003770199 hasPrimaryLocation W30037701991 @default.
• W3003770199 hasRelatedWork W3018421652 @default.
• W3003770199 hasRelatedWork W3021430260 @default.
• W3003770199 hasRelatedWork W3091976719 @default.
• W3003770199 hasRelatedWork W3192840557 @default.
• W3003770199 hasRelatedWork W4220996320 @default.

• next