Matches in SemOpenAlex for { <https://semopenalex.org/work/W2762741128> ?p ?o ?g. }
- W2762741128 endingPage "1272" @default.
- W2762741128 startingPage "1264" @default.
- W2762741128 abstract "Purpose Use adjudication to quantify errors in diabetic retinopathy (DR) grading based on individual graders and majority decision, and to train an improved automated algorithm for DR grading. Design Retrospective analysis. Participants Retinal fundus images from DR screening programs. Methods Images were each graded by the algorithm, U.S. board-certified ophthalmologists, and retinal specialists. The adjudicated consensus of the retinal specialists served as the reference standard. Main Outcome Measures For agreement between different graders as well as between the graders and the algorithm, we measured the (quadratic-weighted) kappa score. To compare the performance of different forms of manual grading and the algorithm for various DR severity cutoffs (e.g., mild or worse DR, moderate or worse DR), we measured area under the curve (AUC), sensitivity, and specificity. Results Of the 193 discrepancies between adjudication by retinal specialists and majority decision of ophthalmologists, the most common were missing microaneurysm (MAs) (36%), artifacts (20%), and misclassified hemorrhages (16%). Relative to the reference standard, the kappa for individual retinal specialists, ophthalmologists, and algorithm ranged from 0.82 to 0.91, 0.80 to 0.84, and 0.84, respectively. For moderate or worse DR, the majority decision of ophthalmologists had a sensitivity of 0.838 and specificity of 0.981. The algorithm had a sensitivity of 0.971, specificity of 0.923, and AUC of 0.986. For mild or worse DR, the algorithm had a sensitivity of 0.970, specificity of 0.917, and AUC of 0.986. By using a small number of adjudicated consensus grades as a tuning dataset and higher-resolution images as input, the algorithm improved in AUC from 0.934 to 0.986 for moderate or worse DR. Conclusions Adjudication reduces the errors in DR grading. A small set of adjudicated DR grades allows substantial improvements in algorithm performance. The resulting algorithm's performance was on par with that of individual U.S. Board-Certified ophthalmologists and retinal specialists. Use adjudication to quantify errors in diabetic retinopathy (DR) grading based on individual graders and majority decision, and to train an improved automated algorithm for DR grading. Retrospective analysis. Retinal fundus images from DR screening programs. Images were each graded by the algorithm, U.S. board-certified ophthalmologists, and retinal specialists. The adjudicated consensus of the retinal specialists served as the reference standard. For agreement between different graders as well as between the graders and the algorithm, we measured the (quadratic-weighted) kappa score. To compare the performance of different forms of manual grading and the algorithm for various DR severity cutoffs (e.g., mild or worse DR, moderate or worse DR), we measured area under the curve (AUC), sensitivity, and specificity. Of the 193 discrepancies between adjudication by retinal specialists and majority decision of ophthalmologists, the most common were missing microaneurysm (MAs) (36%), artifacts (20%), and misclassified hemorrhages (16%). Relative to the reference standard, the kappa for individual retinal specialists, ophthalmologists, and algorithm ranged from 0.82 to 0.91, 0.80 to 0.84, and 0.84, respectively. For moderate or worse DR, the majority decision of ophthalmologists had a sensitivity of 0.838 and specificity of 0.981. The algorithm had a sensitivity of 0.971, specificity of 0.923, and AUC of 0.986. For mild or worse DR, the algorithm had a sensitivity of 0.970, specificity of 0.917, and AUC of 0.986. By using a small number of adjudicated consensus grades as a tuning dataset and higher-resolution images as input, the algorithm improved in AUC from 0.934 to 0.986 for moderate or worse DR. Adjudication reduces the errors in DR grading. A small set of adjudicated DR grades allows substantial improvements in algorithm performance. The resulting algorithm's performance was on par with that of individual U.S. Board-Certified ophthalmologists and retinal specialists." @default.
- W2762741128 created "2017-10-20" @default.
- W2762741128 creator A5036282361 @default.
- W2762741128 creator A5053822927 @default.
- W2762741128 creator A5060000122 @default.
- W2762741128 creator A5060909587 @default.
- W2762741128 creator A5063867870 @default.
- W2762741128 creator A5066927687 @default.
- W2762741128 creator A5076216352 @default.
- W2762741128 creator A5080804221 @default.
- W2762741128 date "2018-08-01" @default.
- W2762741128 modified "2023-10-15" @default.
- W2762741128 title "Grader Variability and the Importance of Reference Standards for Evaluating Machine Learning Models for Diabetic Retinopathy" @default.
- W2762741128 cites W1505156782 @default.
- W2762741128 cites W2004831170 @default.
- W2762741128 cites W2010120422 @default.
- W2762741128 cites W2029023570 @default.
- W2762741128 cites W2053154970 @default.
- W2762741128 cites W2073244572 @default.
- W2762741128 cites W2112796928 @default.
- W2762741128 cites W2115531174 @default.
- W2762741128 cites W2143702645 @default.
- W2762741128 cites W2148309496 @default.
- W2762741128 cites W2149430368 @default.
- W2762741128 cites W2183341477 @default.
- W2762741128 cites W2479953698 @default.
- W2762741128 cites W2557738935 @default.
- W2762741128 cites W2581082771 @default.
- W2762741128 cites W2598442119 @default.
- W2762741128 cites W2772246530 @default.
- W2762741128 cites W2772723798 @default.
- W2762741128 cites W2919115771 @default.
- W2762741128 cites W4240531687 @default.
- W2762741128 doi "https://doi.org/10.1016/j.ophtha.2018.01.034" @default.
- W2762741128 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/29548646" @default.
- W2762741128 hasPublicationYear "2018" @default.
- W2762741128 type Work @default.
- W2762741128 sameAs 2762741128 @default.
- W2762741128 citedByCount "334" @default.
- W2762741128 countsByYear W27627411282017 @default.
- W2762741128 countsByYear W27627411282018 @default.
- W2762741128 countsByYear W27627411282019 @default.
- W2762741128 countsByYear W27627411282020 @default.
- W2762741128 countsByYear W27627411282021 @default.
- W2762741128 countsByYear W27627411282022 @default.
- W2762741128 countsByYear W27627411282023 @default.
- W2762741128 crossrefType "journal-article" @default.
- W2762741128 hasAuthorship W2762741128A5036282361 @default.
- W2762741128 hasAuthorship W2762741128A5053822927 @default.
- W2762741128 hasAuthorship W2762741128A5060000122 @default.
- W2762741128 hasAuthorship W2762741128A5060909587 @default.
- W2762741128 hasAuthorship W2762741128A5063867870 @default.
- W2762741128 hasAuthorship W2762741128A5066927687 @default.
- W2762741128 hasAuthorship W2762741128A5076216352 @default.
- W2762741128 hasAuthorship W2762741128A5080804221 @default.
- W2762741128 hasBestOaLocation W27627411281 @default.
- W2762741128 hasConcept C11413529 @default.
- W2762741128 hasConcept C118487528 @default.
- W2762741128 hasConcept C119767625 @default.
- W2762741128 hasConcept C119857082 @default.
- W2762741128 hasConcept C127413603 @default.
- W2762741128 hasConcept C134018914 @default.
- W2762741128 hasConcept C147176958 @default.
- W2762741128 hasConcept C154945302 @default.
- W2762741128 hasConcept C163864269 @default.
- W2762741128 hasConcept C17744445 @default.
- W2762741128 hasConcept C199539241 @default.
- W2762741128 hasConcept C204434341 @default.
- W2762741128 hasConcept C2524010 @default.
- W2762741128 hasConcept C2776391266 @default.
- W2762741128 hasConcept C2777286243 @default.
- W2762741128 hasConcept C2778724333 @default.
- W2762741128 hasConcept C2779829184 @default.
- W2762741128 hasConcept C2780827179 @default.
- W2762741128 hasConcept C33923547 @default.
- W2762741128 hasConcept C41008148 @default.
- W2762741128 hasConcept C555293320 @default.
- W2762741128 hasConcept C71924100 @default.
- W2762741128 hasConceptScore W2762741128C11413529 @default.
- W2762741128 hasConceptScore W2762741128C118487528 @default.
- W2762741128 hasConceptScore W2762741128C119767625 @default.
- W2762741128 hasConceptScore W2762741128C119857082 @default.
- W2762741128 hasConceptScore W2762741128C127413603 @default.
- W2762741128 hasConceptScore W2762741128C134018914 @default.
- W2762741128 hasConceptScore W2762741128C147176958 @default.
- W2762741128 hasConceptScore W2762741128C154945302 @default.
- W2762741128 hasConceptScore W2762741128C163864269 @default.
- W2762741128 hasConceptScore W2762741128C17744445 @default.
- W2762741128 hasConceptScore W2762741128C199539241 @default.
- W2762741128 hasConceptScore W2762741128C204434341 @default.
- W2762741128 hasConceptScore W2762741128C2524010 @default.
- W2762741128 hasConceptScore W2762741128C2776391266 @default.
- W2762741128 hasConceptScore W2762741128C2777286243 @default.
- W2762741128 hasConceptScore W2762741128C2778724333 @default.
- W2762741128 hasConceptScore W2762741128C2779829184 @default.
- W2762741128 hasConceptScore W2762741128C2780827179 @default.
- W2762741128 hasConceptScore W2762741128C33923547 @default.
- W2762741128 hasConceptScore W2762741128C41008148 @default.