FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4377015596> ?p ?o ?g. }

• W4377015596 endingPage "S607" @default.
• W4377015596 startingPage "S607" @default.
• W4377015596 abstract "With the application of artificial intelligence (AI) in medicine becoming increasingly diverse and the combination of electrocardiogram (ECG) and AI presenting potential clinical implications, the age predicted by AI based on ECG (ECG-age) can be a good predictive measure of cardiovascular diseases and mortality. We developed a convolutional neural network (CNN)-based ECG-age prediction model and investigated whether the age difference (delta age) from CNN-based ECG-age minus chronologic age could predict the future risk of atrial fibrillation (AF) and mortality. An institutional set of 2,647,074 ECG data from 266,283 patients from 2005 to 2020 was screened and two hold-out datasets were created. 1-dimensional CNN was trained to predict patient’s ECG-age using ECGs from the internal training dataset (737,884 ECGs from 195,108 patients). We performed internal validation using the internal test dataset (468,391 ECGs from 48,789 patients) and external validation using four external ECG datasets (6,845 ECGs from external check-ups, 45,610 ECGs from UK Biobank [UKB], 10,198 ECGs from Shaoxing, and 21,152 ECGs from PTB-XL). The delta age was calculated in the internal test and UKB, and participants in each dataset were divided into normal (delta age≤7) or aged (delta age>7) group according to the delta age. The risk of AF and mortality were analyzed according to the delta age group using Kaplan-Meier analysis and multivariate logistic regression. The mean absolute error, reflecting the model's predictability, was 6.81 in the internal test dataset and 4.71, 8.42, 7.08, and 7.79 in the four external datasets, respectively. Compared to normal, the aged group had a lower actual age. The aged group had higher cumulative incidences of new-onset AF and overall mortality in both datasets (p < 0.001, respectively). Moreover, aged group had higher risks of early-onset AF (AF before 66; odds ratio [95% confidence interval], 2.10 [1.84-2.38] in the internal test; 2.10 [1.34-3.17] in the UKB) and early mortality (death before 70; 1.13 [1.02-1.26] in the internal test; 1.77 [1.13-2.68] in the UKB). The ECG-derived delta age showed a statistically significant association with the risk of AF and mortality. The AI-predicted ECG-age can reflect physiological age which might be a potentially useful risk predictor of AF and mortality in general population." @default.
• W4377015596 created "2023-05-19" @default.
• W4377015596 creator A5014897843 @default.
• W4377015596 creator A5050000016 @default.
• W4377015596 creator A5063502620 @default.
• W4377015596 creator A5069907641 @default.
• W4377015596 creator A5071679762 @default.
• W4377015596 creator A5073037699 @default.
• W4377015596 date "2023-05-01" @default.
• W4377015596 modified "2023-09-26" @default.
• W4377015596 title "PO-04-153 CONVOLUTIONAL NEURAL NETWORK-ESTIMATED ELECTROCARDIOGRAPHIC AGE CAN PREDICT THE FUTURE RISK OF ATRIAL FIBRILLATION AND MORTALITY" @default.
• W4377015596 doi "https://doi.org/10.1016/j.hrthm.2023.03.1282" @default.
• W4377015596 hasPublicationYear "2023" @default.
• W4377015596 type Work @default.
• W4377015596 citedByCount "0" @default.
• W4377015596 crossrefType "journal-article" @default.
• W4377015596 hasAuthorship W4377015596A5014897843 @default.
• W4377015596 hasAuthorship W4377015596A5050000016 @default.
• W4377015596 hasAuthorship W4377015596A5063502620 @default.
• W4377015596 hasAuthorship W4377015596A5069907641 @default.
• W4377015596 hasAuthorship W4377015596A5071679762 @default.
• W4377015596 hasAuthorship W4377015596A5073037699 @default.
• W4377015596 hasBestOaLocation W43770155961 @default.
• W4377015596 hasConcept C126322002 @default.
• W4377015596 hasConcept C151956035 @default.
• W4377015596 hasConcept C164705383 @default.
• W4377015596 hasConcept C2779161974 @default.
• W4377015596 hasConcept C2780040984 @default.
• W4377015596 hasConcept C71924100 @default.
• W4377015596 hasConceptScore W4377015596C126322002 @default.
• W4377015596 hasConceptScore W4377015596C151956035 @default.
• W4377015596 hasConceptScore W4377015596C164705383 @default.
• W4377015596 hasConceptScore W4377015596C2779161974 @default.
• W4377015596 hasConceptScore W4377015596C2780040984 @default.
• W4377015596 hasConceptScore W4377015596C71924100 @default.
• W4377015596 hasIssue "5" @default.
• W4377015596 hasLocation W43770155961 @default.
• W4377015596 hasOpenAccess W4377015596 @default.
• W4377015596 hasPrimaryLocation W43770155961 @default.
• W4377015596 hasRelatedWork W1967841144 @default.
• W4377015596 hasRelatedWork W2018480046 @default.
• W4377015596 hasRelatedWork W2064495796 @default.
• W4377015596 hasRelatedWork W2319489406 @default.
• W4377015596 hasRelatedWork W2333279585 @default.
• W4377015596 hasRelatedWork W2603808924 @default.
• W4377015596 hasRelatedWork W2763816051 @default.
• W4377015596 hasRelatedWork W2793053827 @default.
• W4377015596 hasRelatedWork W2889073610 @default.
• W4377015596 hasRelatedWork W3132289493 @default.
• W4377015596 hasVolume "20" @default.
• W4377015596 isParatext "false" @default.
• W4377015596 isRetracted "false" @default.
• W4377015596 workType "article" @default.