FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4384282814> ?p ?o ?g. }

• W4384282814 endingPage "1664" @default.
• W4384282814 startingPage "1657" @default.
• W4384282814 abstract "Abstract Objective To assess large language models on their ability to accurately infer cancer disease response from free-text radiology reports. Materials and Methods We assembled 10 602 computed tomography reports from cancer patients seen at a single institution. All reports were classified into: no evidence of disease, partial response, stable disease, or progressive disease. We applied transformer models, a bidirectional long short-term memory model, a convolutional neural network model, and conventional machine learning methods to this task. Data augmentation using sentence permutation with consistency loss as well as prompt-based fine-tuning were used on the best-performing models. Models were validated on a hold-out test set and an external validation set based on Response Evaluation Criteria in Solid Tumors (RECIST) classifications. Results The best-performing model was the GatorTron transformer which achieved an accuracy of 0.8916 on the test set and 0.8919 on the RECIST validation set. Data augmentation further improved the accuracy to 0.8976. Prompt-based fine-tuning did not further improve accuracy but was able to reduce the number of training reports to 500 while still achieving good performance. Discussion These models could be used by researchers to derive progression-free survival in large datasets. It may also serve as a decision support tool by providing clinicians an automated second opinion of disease response. Conclusions Large clinical language models demonstrate potential to infer cancer disease response from radiology reports at scale. Data augmentation techniques are useful to further improve performance. Prompt-based fine-tuning can significantly reduce the size of the training dataset." @default.
• W4384282814 created "2023-07-15" @default.
• W4384282814 creator A5001400236 @default.
• W4384282814 creator A5004366011 @default.
• W4384282814 creator A5010203164 @default.
• W4384282814 creator A5012390840 @default.
• W4384282814 creator A5014671025 @default.
• W4384282814 creator A5020084287 @default.
• W4384282814 creator A5022140309 @default.
• W4384282814 creator A5023497645 @default.
• W4384282814 creator A5029580390 @default.
• W4384282814 creator A5030515491 @default.
• W4384282814 creator A5041215917 @default.
• W4384282814 creator A5041655333 @default.
• W4384282814 creator A5052752104 @default.
• W4384282814 creator A5056330876 @default.
• W4384282814 creator A5060485249 @default.
• W4384282814 creator A5062344524 @default.
• W4384282814 creator A5069354985 @default.
• W4384282814 creator A5070409002 @default.
• W4384282814 creator A5078925966 @default.
• W4384282814 creator A5084740033 @default.
• W4384282814 creator A5089178032 @default.
• W4384282814 creator A5090494241 @default.
• W4384282814 creator A5091760265 @default.
• W4384282814 date "2023-07-14" @default.
• W4384282814 modified "2023-10-17" @default.
• W4384282814 title "Inferring cancer disease response from radiology reports using large language models with data augmentation and prompting" @default.
• W4384282814 cites W1810662963 @default.
• W4384282814 cites W1971159432 @default.
• W4384282814 cites W2050762896 @default.
• W4384282814 cites W2121274101 @default.
• W4384282814 cites W2345195116 @default.
• W4384282814 cites W2560688864 @default.
• W4384282814 cites W2602748113 @default.
• W4384282814 cites W2786601674 @default.
• W4384282814 cites W2911489562 @default.
• W4384282814 cites W2952455110 @default.
• W4384282814 cites W2963676638 @default.
• W4384282814 cites W2967056613 @default.
• W4384282814 cites W2967365762 @default.
• W4384282814 cites W3047190783 @default.
• W4384282814 cites W3109948322 @default.
• W4384282814 cites W3135550863 @default.
• W4384282814 cites W4200481888 @default.
• W4384282814 cites W4206214875 @default.
• W4384282814 cites W4282983782 @default.
• W4384282814 cites W4308847378 @default.
• W4384282814 cites W4312220150 @default.
• W4384282814 doi "https://doi.org/10.1093/jamia/ocad133" @default.
• W4384282814 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/37451682" @default.
• W4384282814 hasPublicationYear "2023" @default.
• W4384282814 type Work @default.
• W4384282814 citedByCount "0" @default.
• W4384282814 crossrefType "journal-article" @default.
• W4384282814 hasAuthorship W4384282814A5001400236 @default.
• W4384282814 hasAuthorship W4384282814A5004366011 @default.
• W4384282814 hasAuthorship W4384282814A5010203164 @default.
• W4384282814 hasAuthorship W4384282814A5012390840 @default.
• W4384282814 hasAuthorship W4384282814A5014671025 @default.
• W4384282814 hasAuthorship W4384282814A5020084287 @default.
• W4384282814 hasAuthorship W4384282814A5022140309 @default.
• W4384282814 hasAuthorship W4384282814A5023497645 @default.
• W4384282814 hasAuthorship W4384282814A5029580390 @default.
• W4384282814 hasAuthorship W4384282814A5030515491 @default.
• W4384282814 hasAuthorship W4384282814A5041215917 @default.
• W4384282814 hasAuthorship W4384282814A5041655333 @default.
• W4384282814 hasAuthorship W4384282814A5052752104 @default.
• W4384282814 hasAuthorship W4384282814A5056330876 @default.
• W4384282814 hasAuthorship W4384282814A5060485249 @default.
• W4384282814 hasAuthorship W4384282814A5062344524 @default.
• W4384282814 hasAuthorship W4384282814A5069354985 @default.
• W4384282814 hasAuthorship W4384282814A5070409002 @default.
• W4384282814 hasAuthorship W4384282814A5078925966 @default.
• W4384282814 hasAuthorship W4384282814A5084740033 @default.
• W4384282814 hasAuthorship W4384282814A5089178032 @default.
• W4384282814 hasAuthorship W4384282814A5090494241 @default.
• W4384282814 hasAuthorship W4384282814A5091760265 @default.
• W4384282814 hasBestOaLocation W43842828141 @default.
• W4384282814 hasConcept C119857082 @default.
• W4384282814 hasConcept C124101348 @default.
• W4384282814 hasConcept C137293760 @default.
• W4384282814 hasConcept C142724271 @default.
• W4384282814 hasConcept C154945302 @default.
• W4384282814 hasConcept C169903167 @default.
• W4384282814 hasConcept C204321447 @default.
• W4384282814 hasConcept C2778822529 @default.
• W4384282814 hasConcept C2779134260 @default.
• W4384282814 hasConcept C2779984678 @default.
• W4384282814 hasConcept C41008148 @default.
• W4384282814 hasConcept C58489278 @default.
• W4384282814 hasConcept C71924100 @default.
• W4384282814 hasConceptScore W4384282814C119857082 @default.
• W4384282814 hasConceptScore W4384282814C124101348 @default.
• W4384282814 hasConceptScore W4384282814C137293760 @default.
• W4384282814 hasConceptScore W4384282814C142724271 @default.
• W4384282814 hasConceptScore W4384282814C154945302 @default.
• W4384282814 hasConceptScore W4384282814C169903167 @default.

• next