PhD Scientific Days 2024

Cardiovascular Medicine and Research II.

Machine Learning-Based Mortality Prediction in Patients Undergoing CRT Implantation Using the SEMMELWEIS-CRT Score: Validation in the European CRT Survey I Dataset

Előadó neve

Dr. Tokodi, Márton, PhD

Neptun code

zieufc

Előadó munkahelye

Heart and Vascular Center, Semmelweis University, Budapest, Hungary

Előadó telefonszáma

+36304652356

Előadó e-mail címe

tokmarton@gmail.com

Az előadás címe

Machine Learning-Based Mortality Prediction in Patients Undergoing CRT Implantation Using the SEMMELWEIS-CRT Score: Validation in the European CRT Survey I Dataset

Szerző(k) neve és munkahelye

Márton Tokodi¹, Annamária Kosztin¹, Attila Kovács¹, László Gellér¹, Walter Richard Schwertner¹, Boglárka Veres¹, Anett Behon¹, Christiane Lober², Nigussie Bogale³, Cecilia Linde⁴, Camilla Normand⁵, Kenneth Dickstein⁵, Béla Merkely¹

1: Heart and Vascular Center, Semmelweis University, Budapest, Hungary
2: Institut für Herzinfarktforschung, Ludwigshafen, Germany
3: Department of Heart Disease, Haukeland University Hospital, Bergen, Norway
4: Division of Cardiology, Department of Medicine, Karolinska Institutet, Stockholm, Sweden
5: Cardiology Division, Stavanger University Hospital, Stavanger, Norway

Bemutatás módja

Szóbeli

Szekció

Cardiovascular Medicine and Research II.

Language of the presentation

English

Preferred session

Cardiovascular Medicine and Research

Összefoglaló szövege

Introduction: Cardiac resynchronization therapy (CRT) plays a crucial role in the management of heart failure patients. Nevertheless, significant variation in clinical outcomes can still be observed among patients undergoing CRT implantation, which has prompted the development of various machine learning (ML)-based risk stratification tools in this patient population. One such tool is the publicly available SEMMELWEIS-CRT score, which exhibited impressive performance during internal validation but has not been validated externally yet.

Aims: We aimed to externally validate the SEMMELWEIS-CRT score for predicting 1-year all-cause mortality in the European CRT Survey I dataset – a large multi-center cohort of patients undergoing CRT implantation.

Methods: The SEMMELWEIS-CRT score is an ML-based risk stratification tool designed to predict all-cause mortality in patients undergoing CRT implantation. It is a random forest classifier trained in a large retrospective single-center dataset of CRT candidates (n=1,510), taking 33 clinical features as input. In the present study, we applied it to the data of 1,367 patients from the European CRT Survey I dataset.

Results: During the 1-year follow-up period, 122 (9%) of the 1,367 patients died. The SEMMELWEIS-CRT score predicted 1-year mortality with an area under the receiver operating characteristic curve (AUC) of 0.729 [0.682 – 0.776], which concurred with the performance measured during internal validation (AUC: 0.768 [0.674 – 0.861], p=0.466). Moreover, the SEMMELWEIS-CRT score outperformed multiple conventional statistics-based risk scores, including the VALID-CRT (AUC: 0.658 [0.607 – 0.709], p=0.022), the EAARN (AUC: 0.665 [0.619 – 0.711], p=0.016), and the ScREEN scores (AUC: 0.645 [0.600 – 0.691], p=0.003).

Conclusions: In the European CRT Survey I dataset, the SEMMELWEIS-CRT score predicted 1-year all-cause mortality with good discriminatory power, confirming its generalizability and demonstrating its potential clinical utility as an ML-based risk stratification tool in CRT candidates.

Funding: Project no. RRF-2.3.1-21-2022-00004 (MILAB) has been implemented with the support provided by the European Union. Márton Tokodi was supported by the New National Excellence Program (ÚNKP-23-4-II-SE-39) of the Ministry of Culture and Innovation in Hungary from the National Research, Development, and Innovation Fund.

University

Semmelweis University

Supervisor

N/A

Publication of my abstract

I do not give consent to the publication of my abstract on the website of the congress.

Kind

Szabad

Status

elfogadva

Accepted presentation method

szóbeli

Előadás fájl jóváhagyás

nem rendelkezett róla

Előadó

4173

Start

14:45

End

14:55