FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2545581649> ?p ?o ?g. }

• W2545581649 abstract "Abstract Deep learning (DL) has revolutionized the field of computer vision and image processing. In medical imaging, algorithmic solutions based on DL have been shown to achieve high performance on tasks that previously required medical experts. However, DL-based solutions for disease detection have been proposed without methods to quantify and control their uncertainty in a decision. In contrast, a physician knows whether she is uncertain about a case and will consult more experienced colleagues if needed. Here we evaluate drop-out based Bayesian uncertainty measures for DL in diagnosing diabetic retinopathy (DR) from fundus images and show that it captures uncertainty better than straightforward alternatives. Furthermore, we show that uncertainty informed decision referral can improve diagnostic performance. Experiments across different networks, tasks and datasets show robust generalization. Depending on network capacity and task/dataset difficulty, we surpass 85% sensitivity and 80% specificity as recommended by the NHS when referring 0%–20% of the most uncertain decisions for further inspection. We analyse causes of uncertainty by relating intuitions from 2D visualizations to the high-dimensional image space. While uncertainty is sensitive to clinically relevant cases, sensitivity to unfamiliar data samples is task dependent, but can be rendered more robust." @default.
• W2545581649 created "2016-11-04" @default.
• W2545581649 creator A5021786642 @default.
• W2545581649 creator A5043130208 @default.
• W2545581649 creator A5060895798 @default.
• W2545581649 creator A5006418191 @default.
• W2545581649 date "2016-10-28" @default.
• W2545581649 modified "2023-09-26" @default.
• W2545581649 title "Leveraging uncertainty information from deep neural networks for disease detection" @default.
• W2545581649 cites W1494192115 @default.
• W2545581649 cites W1567512734 @default.
• W2545581649 cites W1953868286 @default.
• W2545581649 cites W1965770722 @default.
• W2545581649 cites W1969496006 @default.
• W2545581649 cites W1982366717 @default.
• W2545581649 cites W2017674682 @default.
• W2545581649 cites W2022058268 @default.
• W2545581649 cites W2047229728 @default.
• W2545581649 cites W2069816479 @default.
• W2545581649 cites W2076063813 @default.
• W2545581649 cites W2083780116 @default.
• W2545581649 cites W2088422930 @default.
• W2545581649 cites W2097875638 @default.
• W2545581649 cites W2104787607 @default.
• W2545581649 cites W2111051539 @default.
• W2545581649 cites W2123634291 @default.
• W2545581649 cites W2149430368 @default.
• W2545581649 cites W22040386 @default.
• W2545581649 cites W2233668321 @default.
• W2545581649 cites W2253429366 @default.
• W2545581649 cites W2302517508 @default.
• W2545581649 cites W2344912502 @default.
• W2545581649 cites W2345010043 @default.
• W2545581649 cites W2346062110 @default.
• W2545581649 cites W2401520370 @default.
• W2545581649 cites W2404901863 @default.
• W2545581649 cites W2482922118 @default.
• W2545581649 cites W2502390809 @default.
• W2545581649 cites W2581082771 @default.
• W2545581649 cites W2919115771 @default.
• W2545581649 cites W2963024893 @default.
• W2545581649 cites W3130201740 @default.
• W2545581649 cites W4211049957 @default.
• W2545581649 cites W4246020459 @default.
• W2545581649 doi "https://doi.org/10.1101/084210" @default.
• W2545581649 hasPublicationYear "2016" @default.
• W2545581649 type Work @default.
• W2545581649 sameAs 2545581649 @default.
• W2545581649 citedByCount "4" @default.
• W2545581649 countsByYear W25455816492017 @default.
• W2545581649 countsByYear W25455816492019 @default.
• W2545581649 crossrefType "posted-content" @default.
• W2545581649 hasAuthorship W2545581649A5006418191 @default.
• W2545581649 hasAuthorship W2545581649A5021786642 @default.
• W2545581649 hasAuthorship W2545581649A5043130208 @default.
• W2545581649 hasAuthorship W2545581649A5060895798 @default.
• W2545581649 hasBestOaLocation W25455816491 @default.
• W2545581649 hasConcept C107673813 @default.
• W2545581649 hasConcept C119857082 @default.
• W2545581649 hasConcept C127413603 @default.
• W2545581649 hasConcept C134306372 @default.
• W2545581649 hasConcept C154945302 @default.
• W2545581649 hasConcept C162324750 @default.
• W2545581649 hasConcept C177148314 @default.
• W2545581649 hasConcept C187736073 @default.
• W2545581649 hasConcept C21200559 @default.
• W2545581649 hasConcept C24326235 @default.
• W2545581649 hasConcept C2780451532 @default.
• W2545581649 hasConcept C33923547 @default.
• W2545581649 hasConcept C41008148 @default.
• W2545581649 hasConceptScore W2545581649C107673813 @default.
• W2545581649 hasConceptScore W2545581649C119857082 @default.
• W2545581649 hasConceptScore W2545581649C127413603 @default.
• W2545581649 hasConceptScore W2545581649C134306372 @default.
• W2545581649 hasConceptScore W2545581649C154945302 @default.
• W2545581649 hasConceptScore W2545581649C162324750 @default.
• W2545581649 hasConceptScore W2545581649C177148314 @default.
• W2545581649 hasConceptScore W2545581649C187736073 @default.
• W2545581649 hasConceptScore W2545581649C21200559 @default.
• W2545581649 hasConceptScore W2545581649C24326235 @default.
• W2545581649 hasConceptScore W2545581649C2780451532 @default.
• W2545581649 hasConceptScore W2545581649C33923547 @default.
• W2545581649 hasConceptScore W2545581649C41008148 @default.
• W2545581649 hasLocation W25455816491 @default.
• W2545581649 hasLocation W25455816492 @default.
• W2545581649 hasOpenAccess W2545581649 @default.
• W2545581649 hasPrimaryLocation W25455816491 @default.
• W2545581649 hasRelatedWork W2081647779 @default.
• W2545581649 hasRelatedWork W2961085424 @default.
• W2545581649 hasRelatedWork W2989932438 @default.
• W2545581649 hasRelatedWork W3046775127 @default.
• W2545581649 hasRelatedWork W3209574120 @default.
• W2545581649 hasRelatedWork W4205958290 @default.
• W2545581649 hasRelatedWork W4286629047 @default.
• W2545581649 hasRelatedWork W4306321456 @default.
• W2545581649 hasRelatedWork W4306674287 @default.
• W2545581649 hasRelatedWork W4224009465 @default.
• W2545581649 isParatext "false" @default.
• W2545581649 isRetracted "false" @default.
• W2545581649 magId "2545581649" @default.

• next