Predictive Power for Thrombus Detection after Atrial Appendage Closure: Machine Learning vs. Classical Methods
Device-related thrombus (DRT) after left atrial appendage (LAA) closure is infrequent but correlates with an increased risk of thromboembolism. Therefore, the search for DRT predictors is a topic of interest. In the literature, multivariable methods have been used achieving non-consistent results, a...
| Main Authors: | , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-08-01
|
| Series: | Journal of Personalized Medicine |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-4426/12/9/1413 |
| _version_ | 1797486103434887168 |
|---|---|
| author | Pablo Antúnez-Muiños Víctor Vicente-Palacios Pablo Pérez-Sánchez Jesús Sampedro-Gómez Antonio Sánchez-Puente Pedro Ignacio Dorado-Díaz Luis Nombela-Franco Pablo Salinas Hipólito Gutiérrez-García Ignacio Amat-Santos Vicente Peral Antonio Morcuende Lluis Asmarats Xavier Freixa Ander Regueiro Berenice Caneiro-Queija Rodrigo Estevez-Loureiro Josep Rodés-Cabau Pedro Luis Sánchez Ignacio Cruz-González |
| author_facet | Pablo Antúnez-Muiños Víctor Vicente-Palacios Pablo Pérez-Sánchez Jesús Sampedro-Gómez Antonio Sánchez-Puente Pedro Ignacio Dorado-Díaz Luis Nombela-Franco Pablo Salinas Hipólito Gutiérrez-García Ignacio Amat-Santos Vicente Peral Antonio Morcuende Lluis Asmarats Xavier Freixa Ander Regueiro Berenice Caneiro-Queija Rodrigo Estevez-Loureiro Josep Rodés-Cabau Pedro Luis Sánchez Ignacio Cruz-González |
| author_sort | Pablo Antúnez-Muiños |
| collection | DOAJ |
| description | Device-related thrombus (DRT) after left atrial appendage (LAA) closure is infrequent but correlates with an increased risk of thromboembolism. Therefore, the search for DRT predictors is a topic of interest. In the literature, multivariable methods have been used achieving non-consistent results, and to the best of our knowledge, machine learning techniques have not been used yet for thrombus detection after LAA occlusion. Our aim is to compare both methodologies with respect to predictive power and the search for predictors of DRT. To this end, a multicenter study including 1150 patients who underwent LAA closure was analyzed. Two lines of experiments were performed: with and without resampling. Multivariate and machine learning methodologies were applied to both lines. Predictive power and the extracted predictors for all experiments were gathered. ROC curves of 0.5446 and 0.7974 were obtained for multivariate analysis and machine learning without resampling, respectively. However, the resampling experiment showed no significant difference between them (0.52 vs. 0.53 ROC AUC). A difference between the predictors selected was observed, with the multivariable methodology being more stable. These results question the validity of predictors reported in previous studies and demonstrate their disparity. Furthermore, none of the techniques analyzed is superior to the other for these data. |
| first_indexed | 2024-03-09T23:29:24Z |
| format | Article |
| id | doaj.art-976a175df3f741799f07016d4c42ddf7 |
| institution | Directory Open Access Journal |
| issn | 2075-4426 |
| language | English |
| last_indexed | 2024-03-09T23:29:24Z |
| publishDate | 2022-08-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Journal of Personalized Medicine |
| spelling | doaj.art-976a175df3f741799f07016d4c42ddf72023-11-23T17:12:42ZengMDPI AGJournal of Personalized Medicine2075-44262022-08-01129141310.3390/jpm12091413Predictive Power for Thrombus Detection after Atrial Appendage Closure: Machine Learning vs. Classical MethodsPablo Antúnez-Muiños0Víctor Vicente-Palacios1Pablo Pérez-Sánchez2Jesús Sampedro-Gómez3Antonio Sánchez-Puente4Pedro Ignacio Dorado-Díaz5Luis Nombela-Franco6Pablo Salinas7Hipólito Gutiérrez-García8Ignacio Amat-Santos9Vicente Peral10Antonio Morcuende11Lluis Asmarats12Xavier Freixa13Ander Regueiro14Berenice Caneiro-Queija15Rodrigo Estevez-Loureiro16Josep Rodés-Cabau17Pedro Luis Sánchez18Ignacio Cruz-González19CIBERCV, University Hospital of Salamanca, 37007 Salamanca, SpainPhilips Ibérica, 28050 Madrid, SpainCIBERCV, University Hospital of Salamanca, 37007 Salamanca, SpainCIBERCV, University Hospital of Salamanca, 37007 Salamanca, SpainCIBERCV, University Hospital of Salamanca, 37007 Salamanca, SpainCIBERCV, University Hospital of Salamanca, 37007 Salamanca, SpainInstituto Cardiovascular, Hospital Clínico San Carlos, IdISSC, 28040 Madrid, SpainInstituto Cardiovascular, Hospital Clínico San Carlos, IdISSC, 28040 Madrid, SpainCIBERCV, Instituto de Ciencias del Corazón (ICICOR), Hospital Clínico Universitario de Valladolid, 47003 Valladolid, SpainCIBERCV, Instituto de Ciencias del Corazón (ICICOR), Hospital Clínico Universitario de Valladolid, 47003 Valladolid, SpainDepartment of Cardiology, Health Research Institute of the Balearic Islands (IdISBa), Hospital Universitari Son Espases, 07120 Palma, SpainDepartment of Cardiology, Health Research Institute of the Balearic Islands (IdISBa), Hospital Universitari Son Espases, 07120 Palma, SpainQuebec Heart and Kung Institute, Laval University, Quebec City, QC G1V 0A6, CanadaInstitut Clínic Cardiovascular, Hospital Clínic, Institut d’Investigacions Biomèdiques August Pi I Sunyer (IDIBAPS), 08036 Barcelona, SpainInstitut Clínic Cardiovascular, Hospital Clínic, Institut d’Investigacions Biomèdiques August Pi I Sunyer (IDIBAPS), 08036 Barcelona, SpainUniversity Hospital Alvaro Cunqueiro, 36312 Vigo, SpainUniversity Hospital Alvaro Cunqueiro, 36312 Vigo, SpainQuebec Heart and Kung Institute, Laval University, Quebec City, QC G1V 0A6, CanadaCIBERCV, University Hospital of Salamanca, 37007 Salamanca, SpainCIBERCV, University Hospital of Salamanca, 37007 Salamanca, SpainDevice-related thrombus (DRT) after left atrial appendage (LAA) closure is infrequent but correlates with an increased risk of thromboembolism. Therefore, the search for DRT predictors is a topic of interest. In the literature, multivariable methods have been used achieving non-consistent results, and to the best of our knowledge, machine learning techniques have not been used yet for thrombus detection after LAA occlusion. Our aim is to compare both methodologies with respect to predictive power and the search for predictors of DRT. To this end, a multicenter study including 1150 patients who underwent LAA closure was analyzed. Two lines of experiments were performed: with and without resampling. Multivariate and machine learning methodologies were applied to both lines. Predictive power and the extracted predictors for all experiments were gathered. ROC curves of 0.5446 and 0.7974 were obtained for multivariate analysis and machine learning without resampling, respectively. However, the resampling experiment showed no significant difference between them (0.52 vs. 0.53 ROC AUC). A difference between the predictors selected was observed, with the multivariable methodology being more stable. These results question the validity of predictors reported in previous studies and demonstrate their disparity. Furthermore, none of the techniques analyzed is superior to the other for these data.https://www.mdpi.com/2075-4426/12/9/1413left atrial appendage closuredevice-related thrombosisatrial fibrillationmachine learningmultivariable analysispredictors |
| spellingShingle | Pablo Antúnez-Muiños Víctor Vicente-Palacios Pablo Pérez-Sánchez Jesús Sampedro-Gómez Antonio Sánchez-Puente Pedro Ignacio Dorado-Díaz Luis Nombela-Franco Pablo Salinas Hipólito Gutiérrez-García Ignacio Amat-Santos Vicente Peral Antonio Morcuende Lluis Asmarats Xavier Freixa Ander Regueiro Berenice Caneiro-Queija Rodrigo Estevez-Loureiro Josep Rodés-Cabau Pedro Luis Sánchez Ignacio Cruz-González Predictive Power for Thrombus Detection after Atrial Appendage Closure: Machine Learning vs. Classical Methods Journal of Personalized Medicine left atrial appendage closure device-related thrombosis atrial fibrillation machine learning multivariable analysis predictors |
| title | Predictive Power for Thrombus Detection after Atrial Appendage Closure: Machine Learning vs. Classical Methods |
| title_full | Predictive Power for Thrombus Detection after Atrial Appendage Closure: Machine Learning vs. Classical Methods |
| title_fullStr | Predictive Power for Thrombus Detection after Atrial Appendage Closure: Machine Learning vs. Classical Methods |
| title_full_unstemmed | Predictive Power for Thrombus Detection after Atrial Appendage Closure: Machine Learning vs. Classical Methods |
| title_short | Predictive Power for Thrombus Detection after Atrial Appendage Closure: Machine Learning vs. Classical Methods |
| title_sort | predictive power for thrombus detection after atrial appendage closure machine learning vs classical methods |
| topic | left atrial appendage closure device-related thrombosis atrial fibrillation machine learning multivariable analysis predictors |
| url | https://www.mdpi.com/2075-4426/12/9/1413 |
| work_keys_str_mv | AT pabloantunezmuinos predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT victorvicentepalacios predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT pabloperezsanchez predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT jesussampedrogomez predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT antoniosanchezpuente predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT pedroignaciodoradodiaz predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT luisnombelafranco predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT pablosalinas predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT hipolitogutierrezgarcia predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT ignacioamatsantos predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT vicenteperal predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT antoniomorcuende predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT lluisasmarats predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT xavierfreixa predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT anderregueiro predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT berenicecaneiroqueija predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT rodrigoestevezloureiro predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT joseprodescabau predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT pedroluissanchez predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods AT ignaciocruzgonzalez predictivepowerforthrombusdetectionafteratrialappendageclosuremachinelearningvsclassicalmethods |