Decision Trees for Predicting Mortality in Transcatheter Aortic Valve Implantation

Abstract

Current prognostic risk scores in cardiac surgery do not benefit yet from machine learning (ML). This research aims to create a machine learning model to predict one-year mortality of a patient after transcatheter aortic valve implantation (TAVI). We adopt a modern gradient boosting on decision trees classifier (GBDTs), specifically designed for categorical features. In combination with a recent technique for model interpretations, we developed a feature analysis and selection stage, enabling the identification of the most important features for the prediction. We base our prediction model on the most relevant features, after interpreting and discussing the feature analysis results with clinical experts. We validated our model on 270 consecutive TAVI cases, reaching a C-statistic of 0.83 with CI [0.82, 0.84]. The model has achieved a positive predictive value ranging from 57% to 64%, suggesting that the patient selection made by the heart team of professionals can be further improved by taking into consideration the clinical data we identified as important and by exploiting ML approaches in the development of clinical risk scores. Our approach has shown promising predictive potential also with respect to widespread prognostic risk scores, such as logistic European system for cardiac operative risk evaluation (EuroSCORE II) and the society of thoracic surgeons (STS) risk score, which are broadly adopted by cardiologists worldwide.

Local and Distributed Machine Learning for Inter-hospital Data Utilization: An Application for TAVI Outcome Prediction

“…Some recent studies presented ML models for TAVI outcome prediction. In previous studies, Lopes et al ( 3 , 4 ) developed pipelines for outcome prediction for individual centers. Additionally, Al-Farra et al ( 6 ) and Mamprin et al ( 7 ) showed the accuracy drop on the evaluation of previous traditional risk scores or recent ML models when evaluated on different populations.…”

“…To support patient selection, traditional risk stratification models, either for general cardiac surgery or TAVI specific, are used for mortality estimation (1,2). Other models, exploiting more complex algorithms, have shown higher accuracies when compared to traditional logistic regression-based models (3,4). Nevertheless, mortality estimation models have shown limited prediction accuracy when tested on other center's populations than the one used to generate the models (5)(6)(7)(8).…”

Multiparametric Investigation of Dynamics in Fetal Heart Rate Signals

“…The efforts towards the establishment of a standard technique for the processing and analysis of FHR signals brought the development of different software solutions, methodological approaches, and indicators that could assist the clinical examination of CTG recordings, with particular regard to the FHR signals [ 18 , 19 , 20 , 21 , 22 , 23 , 24 ]. As also happened in the analysis of adult and newborn heart rate signals [ 25 , 26 , 27 , 28 ], most of the newer computerized tools for FHR processing and analysis are based on Artificial Intelligence (AI) algorithms aimed at extracting novel features from the FHR signals, and achieve a more accurate classification of the traces according to the fetal health status [ 29 , 30 , 31 , 32 ]. Among the proposed tools, machine learning algorithms and, in particular, Artificial Neural Networks (ANN) showed promising results in terms of predictability and classification capabilities [ 31 , 32 , 33 , 34 , 35 , 36 ].…”

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

“…Especially when data complexity arises and a large amount of data are involved, ML techniques can outperform conventional regression models [18]. Recently, two ML methods have been developed and published, exploiting gradient boosting on a decision tree algorithm (GBDT) [19,20], to predict one-year mortality for patients treated with TAVI. The GBDT techniques were validated on retrospective patient data from single centers.…”

“…Especially when data complexity arises and a large amount of data are techniques can outperform conventional regression models [18]. Recently, ods have been developed and published, exploiting gradient boosting on algorithm (GBDT) [19,20], to predict one-year mortality for patients treat The GBDT techniques were validated on retrospective patient data from However, external validations were not performed. The research we per manuscript is meant as a natural consequence and continuation of these si searches previously mentioned.…”