Dataset Information

Machine Learning for Predicting Mortality in Transcatheter Aortic Valve Implantation: An Inter-Center Cross Validation Study.

ABSTRACT: Current prognostic risk scores for transcatheter aortic valve implantation (TAVI) do not benefit yet from modern machine learning techniques, which can improve risk stratification of one-year mortality of patients before TAVI. Despite the advancement of machine learning in healthcare, data sharing regulations are very strict and typically prevent exchanging patient data, without the involvement of ethical committees. A very robust validation approach, including 1300 and 631 patients per center, was performed to validate a machine learning model of one center at the other external center with their data, in a mutual fashion. This was achieved without any data exchange but solely by exchanging the models and the data processing pipelines. A dedicated exchange protocol was designed to evaluate and quantify the model's robustness on the data of the external center. Models developed with the larger dataset offered similar or higher prediction accuracy on the external validation. Logistic regression, random forest and CatBoost lead to areas under curve of the ROC of 0.65, 0.67 and 0.65 for the internal validation and of 0.62, 0.66, 0.68 for the external validation, respectively. We propose a scalable exchange protocol which can be further extended on other TAVI centers, but more generally to any other clinical scenario, that could benefit from this validation approach.

SUBMITTER: Mamprin M

PROVIDER: S-EPMC8227005 | biostudies-literature | 2021 Jun

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Machine Learning for Predicting Mortality in Transcatheter Aortic Valve Implantation: An Inter-Center Cross Validation Study.

Mamprin Marco M Lopes Ricardo R RR Zelis Jo M JM Tonino Pim A L PAL van Mourik Martijn S MS Vis Marije M MM Zinger Svitlana S de Mol Bas A J M BAJM de With Peter H N PHN

Journal of cardiovascular development and disease 20210604 6

Current prognostic risk scores for transcatheter aortic valve implantation (TAVI) do not benefit yet from modern machine learning techniques, which can improve risk stratification of one-year mortality of patients before TAVI. Despite the advancement of machine learning in healthcare, data sharing regulations are very strict and typically prevent exchanging patient data, without the involvement of ethical committees. A very robust validation approach, including 1300 and 631 patients per center, ...[more]

PMID: 34199892

Similar Datasets

Project description:ObjectiveTranscatheter aortic valve implantation (TAVI) procedures are increasing rapidly, but the durability of tissue valve and periprocedural complications are not satisfactory. Immune reaction to the galactose-α-1,3 galactose β-1,4-N-acetylglucosamine (α-Gal) and conventional processing protocols of cardiac xenografts lead to calcification. Next-generation TAVI needs to be made with α-Gal-free xenografts by multiple anticalcification therapies to avoid immune rejection and enhance durability, and three-dimensional (3D) printing technology to improve the procedural safety.MethodsPorcine pericardia were decellularized and immunologically modified with α-galactosidase. The pericardia were treated by space filler, crosslinked with glutaraldehyde in organic solvent, and detoxified. The sheep-specific nitinol (nickel-titanium memory alloy) wire backbone was made from a 3D-printed model for ovine aortic root. After it passed the fitting test, we manufactured a self-expandable stented valve with the porcine pericardia mounted on the customized nitinol wire-based stent. After in vitro circulation using customized silicone aortic root, we performed TAVI in 9 sheep and obtained hemodynamic, radiological, immunohistopathological, and biochemical results.ResultsThe valve functioned well, with excellent stent fitting and good coronary flow under in vitro circulation. Sheep were sequentially scheduled to be humanely killed until 238 days after TAVI. Echocardiography and cardiac catheterization demonstrated good hemodynamic status and function of the aortic valve. The xenografts were well preserved without α-Gal immune reaction or calcification based on the immunological, radiographic, microscopic, and biochemical examinations.ConclusionsWe proved preclinical safety and efficacy for next-generation α-Gal-free TAVI with multiple anticalcification therapies and 3D-printing technology. A future clinical study is warranted based on these promising preclinical results.

Dataset Information

Machine Learning for Predicting Mortality in Transcatheter Aortic Valve Implantation: An Inter-Center Cross Validation Study.

Publications

Machine Learning for Predicting Mortality in Transcatheter Aortic Valve Implantation: An Inter-Center Cross Validation Study.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets