Bayesian parametric models for survival prediction in medical applications

BMC Medical Research Methodology

Research

Kristy K. Brock¹  Jessica Albuquerque Marques Silva²  Yuan-Mao Lin²  Iwan Paolucci²  Bruno C. Odisio² 

[1] Department of Imaging Physics, The University of Texas MD Anderson Cancer Center, Houston, TX, USA;Department of Interventional Radiology, The University of Texas MD Anderson Cancer Center, Houston, TX, USA;

关键词: Bayesian;    Neural network;    Machine learning;    Survival analysis;    Survival outcome;   

DOI  :  10.1186/s12874-023-02059-4

 received in 2023-04-24, accepted in 2023-10-06,  发布年份 2023

来源: Springer

【 摘 要 】

BackgroundEvidence-based treatment decisions in medicine are made founded on population-level evidence obtained during randomized clinical trials. In an era of personalized medicine, these decisions should be based on the predicted benefit of a treatment on a patient-level. Survival prediction models play a central role as they incorporate the time-to-event and censoring. In medical applications uncertainty is critical especially when treatments differ in their side effect profiles or costs. Additionally, models must be adapted to local populations without diminishing performance and often without the original training data available due to privacy concern. Both points are supported by Bayesian models—yet they are rarely used. The aim of this work is to evaluate Bayesian parametric survival models on public datasets including cardiology, infectious diseases, and oncology.Materials and methodsBayesian parametric survival models based on the Exponential and Weibull distribution were implemented as a Python package. A linear combination and a neural network were used for predicting the parameters of the distributions. A superiority design was used to assess whether Bayesian models are better than commonly used models such as Cox Proportional Hazards, Random Survival Forest, and Neural Network-based Cox Proportional Hazards. In a secondary analysis, overfitting was compared between these models. An equivalence design was used to assess whether the prediction performance of Bayesian models after model updating using Bayes rule is equivalent to retraining on the full dataset.ResultsIn this study, we found that Bayesian parametric survival models perform as good as state-of-the art models while requiring less hyperparameters to be tuned and providing a measure of the uncertainty of the predictions. In addition, these models were less prone to overfitting. Furthermore, we show that updating these models using Bayes rule yields equivalent performance compared to models trained on combined original and new datasets.ConclusionsBayesian parametric survival models are non-inferior to conventional survival models while requiring less hyperparameter tuning, being less prone to overfitting, and allowing model updating using Bayes rule. Further, the Bayesian models provide a measure of the uncertainty on the statistical inference, and, in particular, on the prediction.

【 授权许可】

CC BY   
© The Author(s) 2023

【 图 表 】

Fig. 10

12951_2017_255_Article_IEq50.gif

12951_2016_246_Article_IEq8.gif

12951_2015_155_Article_IEq69.gif

12936_2015_836_Article_IEq14.gif

Fig. 5

Fig. 6

12936_2017_2045_Article_IEq3.gif

Fig. 3

Fig. 3

Fig. 2