Matches in SemOpenAlex for { <https://semopenalex.org/work/W4238734820> ?p ?o ?g. }
- W4238734820 abstract "Abstract Pathologists can label pathologies differently, making it challenging to yield consistent assessments in the absence of one ground truth. To address this problem, we present a DL approach that draws on a cohort of experts, weighs each contribution, and is robust to noisy labels. We collected 100,495 annotations on 20,099 candidate amyloid beta neuropathologies (cerebral amyloid angiopathy (CAA), and cored and diffuse plaques) from three institutions, independently annotated by five experts. DL methods trained on a consensus-of-two strategy yielded 12.6-26% improvements by area under the precision recall curve (AUPRC) when compared to those that learned individualized annotations. This strategy surpassed individual-expert models, even when unfairly assessed on benchmarks favoring them. Moreover, ensembling over individual models was robust to hidden random annotators. In blind prospective tests of 52,555 subsequent expert-annotated images, the models labeled pathologies like their human counterparts (consensus model AUPRC=0.74 cored; 0.69 CAA). This study demonstrates a means to combine multiple ground truths into a common-ground DL model that yields consistent diagnoses informed by multiple and potentially variable expert opinion." @default.
- W4238734820 created "2022-05-12" @default.
- W4238734820 creator A5000414426 @default.
- W4238734820 creator A5001435363 @default.
- W4238734820 creator A5002496711 @default.
- W4238734820 creator A5004078745 @default.
- W4238734820 creator A5018266790 @default.
- W4238734820 creator A5024767343 @default.
- W4238734820 creator A5034847199 @default.
- W4238734820 creator A5036068985 @default.
- W4238734820 creator A5037987993 @default.
- W4238734820 creator A5047367774 @default.
- W4238734820 creator A5088443428 @default.
- W4238734820 creator A5020949783 @default.
- W4238734820 creator A5077567139 @default.
- W4238734820 date "2021-03-12" @default.
- W4238734820 modified "2023-10-10" @default.
- W4238734820 title "Deep learning from multiple experts improves identification of amyloid neuropathologies" @default.
- W4238734820 cites W1942214758 @default.
- W4238734820 cites W1972675781 @default.
- W4238734820 cites W1978293591 @default.
- W4238734820 cites W2010135967 @default.
- W4238734820 cites W2012215271 @default.
- W4238734820 cites W2033917523 @default.
- W4238734820 cites W2034893190 @default.
- W4238734820 cites W2053675728 @default.
- W4238734820 cites W2060132087 @default.
- W4238734820 cites W2061820841 @default.
- W4238734820 cites W2087529125 @default.
- W4238734820 cites W2103243046 @default.
- W4238734820 cites W2112796928 @default.
- W4238734820 cites W2115564009 @default.
- W4238734820 cites W2128228199 @default.
- W4238734820 cites W2129112648 @default.
- W4238734820 cites W2132131059 @default.
- W4238734820 cites W2133665775 @default.
- W4238734820 cites W2150920454 @default.
- W4238734820 cites W2312439098 @default.
- W4238734820 cites W2463189234 @default.
- W4238734820 cites W2524083015 @default.
- W4238734820 cites W2533800772 @default.
- W4238734820 cites W2570634615 @default.
- W4238734820 cites W2581082771 @default.
- W4238734820 cites W2767410506 @default.
- W4238734820 cites W2900743800 @default.
- W4238734820 cites W2905483812 @default.
- W4238734820 cites W2914546793 @default.
- W4238734820 cites W2924530200 @default.
- W4238734820 cites W2940900531 @default.
- W4238734820 cites W2947792531 @default.
- W4238734820 cites W2947825023 @default.
- W4238734820 cites W2956228567 @default.
- W4238734820 cites W2962858109 @default.
- W4238734820 cites W2963772355 @default.
- W4238734820 cites W3020814630 @default.
- W4238734820 cites W3195555725 @default.
- W4238734820 doi "https://doi.org/10.1101/2021.03.12.435050" @default.
- W4238734820 hasPublicationYear "2021" @default.
- W4238734820 type Work @default.
- W4238734820 citedByCount "0" @default.
- W4238734820 crossrefType "posted-content" @default.
- W4238734820 hasAuthorship W4238734820A5000414426 @default.
- W4238734820 hasAuthorship W4238734820A5001435363 @default.
- W4238734820 hasAuthorship W4238734820A5002496711 @default.
- W4238734820 hasAuthorship W4238734820A5004078745 @default.
- W4238734820 hasAuthorship W4238734820A5018266790 @default.
- W4238734820 hasAuthorship W4238734820A5020949783 @default.
- W4238734820 hasAuthorship W4238734820A5024767343 @default.
- W4238734820 hasAuthorship W4238734820A5034847199 @default.
- W4238734820 hasAuthorship W4238734820A5036068985 @default.
- W4238734820 hasAuthorship W4238734820A5037987993 @default.
- W4238734820 hasAuthorship W4238734820A5047367774 @default.
- W4238734820 hasAuthorship W4238734820A5077567139 @default.
- W4238734820 hasAuthorship W4238734820A5088443428 @default.
- W4238734820 hasBestOaLocation W42387348201 @default.
- W4238734820 hasConcept C100660578 @default.
- W4238734820 hasConcept C116834253 @default.
- W4238734820 hasConcept C119857082 @default.
- W4238734820 hasConcept C142724271 @default.
- W4238734820 hasConcept C146849305 @default.
- W4238734820 hasConcept C153180895 @default.
- W4238734820 hasConcept C154945302 @default.
- W4238734820 hasConcept C15744967 @default.
- W4238734820 hasConcept C180747234 @default.
- W4238734820 hasConcept C204321447 @default.
- W4238734820 hasConcept C2777790613 @default.
- W4238734820 hasConcept C2779134260 @default.
- W4238734820 hasConcept C2779483572 @default.
- W4238734820 hasConcept C41008148 @default.
- W4238734820 hasConcept C534262118 @default.
- W4238734820 hasConcept C59822182 @default.
- W4238734820 hasConcept C71924100 @default.
- W4238734820 hasConcept C86803240 @default.
- W4238734820 hasConceptScore W4238734820C100660578 @default.
- W4238734820 hasConceptScore W4238734820C116834253 @default.
- W4238734820 hasConceptScore W4238734820C119857082 @default.
- W4238734820 hasConceptScore W4238734820C142724271 @default.
- W4238734820 hasConceptScore W4238734820C146849305 @default.
- W4238734820 hasConceptScore W4238734820C153180895 @default.
- W4238734820 hasConceptScore W4238734820C154945302 @default.