FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4386906231> ?p ?o ?g. }

• W4386906231 endingPage "S1" @default.
• W4386906231 startingPage "S1" @default.
• W4386906231 abstract "Fatal Coronary Heart Disease (FCHD) is a proxy for sudden cardiac death (SCD), affecting >17 million people/year globally, with high prevalence even in younger adults. Electrocardiographic Artificial Intelligence (ECG-AI) models assessing cardiovascular disease risk have been developed but few exist for FCHD. ECG-AI based FCHD risk prediction tools, which are currently lacking, based on single lead ECG data from wearable devices would enable scalable screening and/or monitoring of younger populations. To develop a single lead ECG-based deep learning model for FCHD risk prediction and assess prediction concordance between paired clinical and Apple Watch ECGs. A FCHD lead I ECG-based 1D convolutional neural network model was developed using 420,834 ECGs (68,094 patients), obtained from UTHSC, TN. The final model was tested on paired lead I clinical and Apple Watch ECGs collected at the same visit from 55 volunteers (unknown FCHD status). The concordance of FCHD risk between the two ECG modalities was assessed using paired correlation analysis. The UTHSC lead I ECG-AI model was developed on 80% data (5-fold cross-validation) and resulted in AUC=0.74 on the 20% holdout data. When implemented on 55 paired ECGs, there was strong positive correlation (Fig. 1) between predictions obtained from the two ECG modalities (correlation coefficient=0.82, p<0.001). FCHD risk may be predicted from 1 lead ECGs using Apple Watches with moderate accuracy and be concordant with clinical ECGs. ECG-AI models implemented on Apple Watch ECGs can help screen large populations for FCHD with ease and at low cost. Adding clinical variables in deep survival models might improve accuracy (further work). Some limitations of this pilot work include: lead I ECG-AI FCHD model is preliminary and not optimized, and the resolution (signal depth) of Apple Watch ECGs needs to be adjusted to match clinical ECGs (A/D conversion), yet such documentation is lacking." @default.
• W4386906231 created "2023-09-21" @default.
• W4386906231 creator A5027592744 @default.
• W4386906231 creator A5047770817 @default.
• W4386906231 creator A5049425464 @default.
• W4386906231 creator A5052539701 @default.
• W4386906231 creator A5062032724 @default.
• W4386906231 creator A5066509740 @default.
• W4386906231 creator A5066607244 @default.
• W4386906231 creator A5071844909 @default.
• W4386906231 date "2023-10-01" @default.
• W4386906231 modified "2023-10-16" @default.
• W4386906231 title "FEASIBILITY OF REMOTE MONITORING FOR FATAL CORONARY HEART DISEASE FROM SINGLE LEAD ECG" @default.
• W4386906231 doi "https://doi.org/10.1016/j.cvdhj.2023.08.002" @default.
• W4386906231 hasPublicationYear "2023" @default.
• W4386906231 type Work @default.
• W4386906231 citedByCount "0" @default.
• W4386906231 crossrefType "journal-article" @default.
• W4386906231 hasAuthorship W4386906231A5027592744 @default.
• W4386906231 hasAuthorship W4386906231A5047770817 @default.
• W4386906231 hasAuthorship W4386906231A5049425464 @default.
• W4386906231 hasAuthorship W4386906231A5052539701 @default.
• W4386906231 hasAuthorship W4386906231A5062032724 @default.
• W4386906231 hasAuthorship W4386906231A5066509740 @default.
• W4386906231 hasAuthorship W4386906231A5066607244 @default.
• W4386906231 hasAuthorship W4386906231A5071844909 @default.
• W4386906231 hasBestOaLocation W43869062311 @default.
• W4386906231 hasConcept C117220453 @default.
• W4386906231 hasConcept C119857082 @default.
• W4386906231 hasConcept C126322002 @default.
• W4386906231 hasConcept C160798450 @default.
• W4386906231 hasConcept C164705383 @default.
• W4386906231 hasConcept C2524010 @default.
• W4386906231 hasConcept C2780040984 @default.
• W4386906231 hasConcept C33923547 @default.
• W4386906231 hasConcept C41008148 @default.
• W4386906231 hasConcept C71924100 @default.
• W4386906231 hasConceptScore W4386906231C117220453 @default.
• W4386906231 hasConceptScore W4386906231C119857082 @default.
• W4386906231 hasConceptScore W4386906231C126322002 @default.
• W4386906231 hasConceptScore W4386906231C160798450 @default.
• W4386906231 hasConceptScore W4386906231C164705383 @default.
• W4386906231 hasConceptScore W4386906231C2524010 @default.
• W4386906231 hasConceptScore W4386906231C2780040984 @default.
• W4386906231 hasConceptScore W4386906231C33923547 @default.
• W4386906231 hasConceptScore W4386906231C41008148 @default.
• W4386906231 hasConceptScore W4386906231C71924100 @default.
• W4386906231 hasIssue "5" @default.
• W4386906231 hasLocation W43869062311 @default.
• W4386906231 hasOpenAccess W4386906231 @default.
• W4386906231 hasPrimaryLocation W43869062311 @default.
• W4386906231 hasRelatedWork W1579712870 @default.
• W4386906231 hasRelatedWork W165209850 @default.
• W4386906231 hasRelatedWork W2008295798 @default.
• W4386906231 hasRelatedWork W2055495421 @default.
• W4386906231 hasRelatedWork W2077766661 @default.
• W4386906231 hasRelatedWork W2252916976 @default.
• W4386906231 hasRelatedWork W2416422041 @default.
• W4386906231 hasRelatedWork W4238670373 @default.
• W4386906231 hasRelatedWork W3112149275 @default.
• W4386906231 hasRelatedWork W3157222668 @default.
• W4386906231 hasVolume "4" @default.
• W4386906231 isParatext "false" @default.
• W4386906231 isRetracted "false" @default.
• W4386906231 workType "article" @default.