Matches in SemOpenAlex for { <https://semopenalex.org/work/W3150641585> ?p ?o ?g. }
- W3150641585 endingPage "1148" @default.
- W3150641585 startingPage "1135" @default.
- W3150641585 abstract "Abstract Objective Reticular pseudodrusen (RPD), a key feature of age-related macular degeneration (AMD), are poorly detected by human experts on standard color fundus photography (CFP) and typically require advanced imaging modalities such as fundus autofluorescence (FAF). The objective was to develop and evaluate the performance of a novel multimodal, multitask, multiattention (M3) deep learning framework on RPD detection. Materials and Methods A deep learning framework (M3) was developed to detect RPD presence accurately using CFP alone, FAF alone, or both, employing >8000 CFP-FAF image pairs obtained prospectively (Age-Related Eye Disease Study 2). The M3 framework includes multimodal (detection from single or multiple image modalities), multitask (training different tasks simultaneously to improve generalizability), and multiattention (improving ensembled feature representation) operation. Performance on RPD detection was compared with state-of-the-art deep learning models and 13 ophthalmologists; performance on detection of 2 other AMD features (geographic atrophy and pigmentary abnormalities) was also evaluated. Results For RPD detection, M3 achieved an area under the receiver-operating characteristic curve (AUROC) of 0.832, 0.931, and 0.933 for CFP alone, FAF alone, and both, respectively. M3 performance on CFP was very substantially superior to human retinal specialists (median F1 score = 0.644 vs 0.350). External validation (the Rotterdam Study) demonstrated high accuracy on CFP alone (AUROC, 0.965). The M3 framework also accurately detected geographic atrophy and pigmentary abnormalities (AUROC, 0.909 and 0.912, respectively), demonstrating its generalizability. Conclusions This study demonstrates the successful development, robust evaluation, and external validation of a novel deep learning framework that enables accessible, accurate, and automated AMD diagnosis and prognosis." @default.
- W3150641585 created "2021-04-13" @default.
- W3150641585 creator A5017028738 @default.
- W3150641585 creator A5022184805 @default.
- W3150641585 creator A5029529953 @default.
- W3150641585 creator A5036302697 @default.
- W3150641585 creator A5039817860 @default.
- W3150641585 creator A5039913574 @default.
- W3150641585 creator A5042874172 @default.
- W3150641585 creator A5048293450 @default.
- W3150641585 creator A5049198875 @default.
- W3150641585 creator A5049602513 @default.
- W3150641585 creator A5052248382 @default.
- W3150641585 creator A5068185303 @default.
- W3150641585 creator A5070251078 @default.
- W3150641585 creator A5083081872 @default.
- W3150641585 creator A5085113833 @default.
- W3150641585 creator A5088588454 @default.
- W3150641585 creator A5090802305 @default.
- W3150641585 date "2021-04-01" @default.
- W3150641585 modified "2023-10-04" @default.
- W3150641585 title "Multimodal, multitask, multiattention (M3) deep learning detection of reticular pseudodrusen: Toward automated and accessible classification of age-related macular degeneration" @default.
- W3150641585 cites W159419684 @default.
- W3150641585 cites W1789824814 @default.
- W3150641585 cites W1796224740 @default.
- W3150641585 cites W1965478508 @default.
- W3150641585 cites W1972642682 @default.
- W3150641585 cites W1974599215 @default.
- W3150641585 cites W1977181227 @default.
- W3150641585 cites W1987084100 @default.
- W3150641585 cites W1990638682 @default.
- W3150641585 cites W2071720755 @default.
- W3150641585 cites W2090783890 @default.
- W3150641585 cites W2090949500 @default.
- W3150641585 cites W2096666997 @default.
- W3150641585 cites W2126598020 @default.
- W3150641585 cites W2129892228 @default.
- W3150641585 cites W2154864949 @default.
- W3150641585 cites W2306170388 @default.
- W3150641585 cites W2320938042 @default.
- W3150641585 cites W2419255724 @default.
- W3150641585 cites W2533800772 @default.
- W3150641585 cites W2575549046 @default.
- W3150641585 cites W2592929672 @default.
- W3150641585 cites W2598520882 @default.
- W3150641585 cites W2602101491 @default.
- W3150641585 cites W2758333670 @default.
- W3150641585 cites W2760607312 @default.
- W3150641585 cites W2765808721 @default.
- W3150641585 cites W2767290858 @default.
- W3150641585 cites W2790299667 @default.
- W3150641585 cites W2796809202 @default.
- W3150641585 cites W2807327270 @default.
- W3150641585 cites W2809814422 @default.
- W3150641585 cites W2888424632 @default.
- W3150641585 cites W2891443166 @default.
- W3150641585 cites W2893458693 @default.
- W3150641585 cites W2896202701 @default.
- W3150641585 cites W2896734556 @default.
- W3150641585 cites W2898192966 @default.
- W3150641585 cites W2901158775 @default.
- W3150641585 cites W2907760128 @default.
- W3150641585 cites W2909356997 @default.
- W3150641585 cites W2917870699 @default.
- W3150641585 cites W2949127080 @default.
- W3150641585 cites W2965133071 @default.
- W3150641585 cites W2974888503 @default.
- W3150641585 cites W3000187180 @default.
- W3150641585 cites W3099733385 @default.
- W3150641585 cites W4242323807 @default.
- W3150641585 cites W4321509030 @default.
- W3150641585 doi "https://doi.org/10.1093/jamia/ocaa302" @default.
- W3150641585 hasPubMedCentralId "https://www.ncbi.nlm.nih.gov/pmc/articles/8200273" @default.
- W3150641585 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/33792724" @default.
- W3150641585 hasPublicationYear "2021" @default.
- W3150641585 type Work @default.
- W3150641585 sameAs 3150641585 @default.
- W3150641585 citedByCount "11" @default.
- W3150641585 countsByYear W31506415852021 @default.
- W3150641585 countsByYear W31506415852022 @default.
- W3150641585 countsByYear W31506415852023 @default.
- W3150641585 crossrefType "journal-article" @default.
- W3150641585 hasAuthorship W3150641585A5017028738 @default.
- W3150641585 hasAuthorship W3150641585A5022184805 @default.
- W3150641585 hasAuthorship W3150641585A5029529953 @default.
- W3150641585 hasAuthorship W3150641585A5036302697 @default.
- W3150641585 hasAuthorship W3150641585A5039817860 @default.
- W3150641585 hasAuthorship W3150641585A5039913574 @default.
- W3150641585 hasAuthorship W3150641585A5042874172 @default.
- W3150641585 hasAuthorship W3150641585A5048293450 @default.
- W3150641585 hasAuthorship W3150641585A5049198875 @default.
- W3150641585 hasAuthorship W3150641585A5049602513 @default.
- W3150641585 hasAuthorship W3150641585A5052248382 @default.
- W3150641585 hasAuthorship W3150641585A5068185303 @default.
- W3150641585 hasAuthorship W3150641585A5070251078 @default.
- W3150641585 hasAuthorship W3150641585A5083081872 @default.
- W3150641585 hasAuthorship W3150641585A5085113833 @default.
- W3150641585 hasAuthorship W3150641585A5088588454 @default.